Download 5. F M F

Central Bank of the Republic of Turkey
5. Financial Markets and Financial Intermediation
In the last quarter, financial markets remained volatile due to international developments. In this
period, the uncertainty over the Fed’s monetary policy continued to be less acute than the previous
year, which fueled the risk appetite; yet the weak global growth outlook affected financial markets
adversely. In addition, growing geopolitical tensions were another factor feeding market volatility. On
the other hand, the decline in raw material prices, oil prices in particular, appears to have had a
favorable impact on the risk premium of countries, such as Turkey, that are dependent on external
energy.
The FCI for Turkey, which is calculated as the weighted average of various financial indicators,
pointed to a neutral reading (close to the historical average) for the third quarter (Chart 5.1). In this
period, each variable in the index mostly provided near-neutral contributions to financial conditions as
well (Chart 5.2). Thanks to the CBRT’s prudent monetary policy stance, the benchmark rate and the
slope of the yield curve continued to help tighten financial conditions in the third quarter, albeit to a
lesser degree. On the other hand, credit conditions were relatively accommodative. The contribution
from loan rates was negative in the first two quarters of 2014, but turned neutral in the third quarter.
Meanwhile, loan standards surpassed their historical average in the third quarter, which stemmed from
the improvement in commercial loan standards. The ongoing decline in loan rates and the course of
loan standards suggest that the downtrend in annual loan growth will no longer continue in the
upcoming period. Yet, the rise in precautionary savings against geopolitical tensions and the CBRT’s
tight liquidity stance may restrain financial easing.
Chart 5.1.
Chart 5.2.
accommodative
Financial Conditions and Credit Growth*
Contributions to FCI**
FCI (standardized)
EMBI
Slope of the Yield Curve
Change in Credits/GDP (right axis)
Loan Rate
Exhange Rate
Benchmark Rate
Loan Standards
Capital Flows
Stock Return
3
18
15
2
3
3
2
2
1
1
0
0
-1
-1
0
-2
-2
-3
-3
12
1
9
0
tightening
6
-1
3
-2
-3
123412341234123412341234123
2008
2009
2010
2011
2012
2013 2014
-3
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3
2008
2009
2010
2011
2012
2013
2014
* For further details on measuring FCI, see the CBT Research Notes in Economics No. 12/31.
** Slope of the yield curve is denoted by the spread between 10-year and 2-year interest rate. For pre-2010 period, 10-year cross currency swap rate is used
due to absence of 10-year interest rates.
Source: CBRT.
5.1. Financial Markets
Global Risk Perceptions
The uncertainty over global monetary policies continued into the third quarter of 2014, with the
Fed ending its bond-buying program and announcing that it would keep policy rates low for a while.
Upward revisions to the second-quarter US growth rate in August and September and the recovery in
Inflation Report 2014-IV
49
Central Bank of the Republic of Turkey
both economic activity and leading indicators spurred the belief that the Fed would deliver a soonerthan-expected policy rate hike. On the other hand, the hints from some board members in October
that the Fed would continue to support economic growth in the face of weakening global growth
boosted the global risk appetite. Meanwhile, amid weak economic activity in the EU and the growing
risk of deflation, the ECB cut the policy rate and the negative interest rate imposed on banks for their
deposits by 10 basis points each and decided to launch a two-year asset-backed securities purchase
program starting in the final quarter of this year. Another highlight of this quarter was the sharp drop in
Japan’s second-quarter growth data, which fueled market expectations that Japan would announce
new stimulus programs. The uncertainty over global monetary policies and the growing geopolitical
risks caused volatility indices to soar in this period (Charts 5.1.1 and 5.1.2). Following the deteriorated risk
sentiment in the beginning of the quarter, global investors fled increasingly towards safe-haven assets
of the US bond market, which caused US bond yields to fall. The recent shift from stocks to government
bonds has added to the decline in US yields (Chart 5.1.1).
Chart 5.1.1.
Chart 5.1.2.
10-Year US Treasury Bond Rates and MOVE Index
JPMVXYEM Volatility Index
(Percent)
3.5
3.0
120
14
14
110
13
13
12
12
11
11
10
10
80
9
9
70
8
8
7
7
6
6
5
5
4
4
100
90
2.5
2.0
60
1.5
50
1014
0714
0414
0114
1013
0713
0413
40
0113
1.0
0113
0213
0313
0413
0513
0613
0713
0813
0913
1013
1113
1213
0114
0214
0314
0414
0514
0614
0714
0814
0914
1014
10-Year USD Treasury Bond Rates (percent)
MOVE Index (right axis)
Source: Bloomberg.
In the third quarter, the economic slowdown across Eastern Europe and Latin America,
particularly after the Fed’s signals, affected the risk sentiment for emerging economies negatively,
leading to an increase in the risk premiums of these countries. The fall in CDS premiums of emerging
economies amid reduced global uncertainty in the first half of 2014 was reversed in the second half
due to the uncertainty about the Fed’s policy rate hike and the geopolitical risks arising from
developments in Iraq and Syria. However, CDS premiums have declined slightly following the growing
global risk appetite amid the Fed’s recent announcements supportive of economic growth.
Meanwhile, Turkey’s sovereign risk premium rose in line with other emerging economies in this period
(Charts 5.1.3 and 5.1.4).
50
Inflation Report 2014-IV
Central Bank of the Republic of Turkey
Chart 5.1.3.
Chart 5.1.4.
Regional EMBI Indices
Changes in CDS*
(Basis Points)
(Basis Points)
EMBI Europe
EMBI Turkey
EMBI Asia
EMBI Latin America
Turkey
Emerging Economies
Selected Emerging Economies
500
500
450
450
0
400
400
-20
-20
350
350
-40
-40
300
300
-60
-60
250
250
-80
-80
200
200
-100
-100
150
150
-120
-120
100
100
-140
-140
20
20
1112
1212
0113
0213
0313
0413
0513
0613
0713
0813
0913
1013
1113
1213
0114
0214
0314
0414
0514
0614
0714
0814
0914
1014
0114
0214
0214
0314
0314
0414
0414
0514
0514
0614
0614
0714
0714
0814
0814
0914
0914
1014
1014
0
* Emerging economies include Brazil, Chile, Colombia, Czech Republic, Hungary, Indonesia, Mexico, Poland, Romania and South Africa. Selected emerging
economies are Brazil, Indonesia and South Africa. Denote changes since January 24.
Source: Bloomberg.
Portfolio Flows
The loss of momentum in growth performances across emerging economies and the
uncertainties surrounding global monetary policies caused emerging market capital flows to follow a
weak and volatile course in the third quarter of 2014 (Chart 5.1.5). Capital flows to Turkey were also
weak and volatile in this period because of increased sovereign risk premiums, and witnessed outflows
despite capital inflows during June-July, causing cumulative portfolio flows from the start of 2014 to stay
at near-zero (Chart 5.1.6).
Chart 5.1.5.
Chart 5.1.6.
Portfolio Flows to Emerging Economies*
Portfolio Flows to Turkey*
(Billion USD)
(Billion USD)
80
80
60
60
2013
2008-2013
Average
40
40
20
20
0
0
25
25
20
20
15
15
10
-20
2008-2013
Average
2013
10
5
5
0
0
-20
2014
-40
December
November
October
September
July
-5
August
May
April
March
June
2014
-5
February
December
November
October
September
July
August
May
June
April
March
February
-60
January
-60
January
-40
* Includes equity and bond flows.
Source: EPFR.
Exchange Rates
The economic recovery signaled by the promising US growth and employment data and
leading indicators were interpreted to mean that the Fed would start a sooner-than-expected
normalization in monetary policy. This possibility caused the DXY index to soar sharply in the third
quarter amid expectations of continued monetary easing in the Euro Area and Japan launching new
quantitative easing measures to support the economy. Meanwhile, emerging market currencies have
Inflation Report 2014-IV
51
Central Bank of the Republic of Turkey
depreciated against the USD since the July reporting period due to the reduced global risk appetite
and the heightening sovereign risk premiums (Chart 5.1.7). The global risk appetite recovered to some
extent following the positive signals from the Fed in October, helping to compensate for some of these
losses. The Turkish lira moved in tandem with the currencies of peer emerging market economies in the
third quarter. The strong relationship between the currency basket and the risk premium continued in
this period and the currency basket increased in line with the rising risk premium. After hitting 2.46 on
July 24 when the July Inflation Report was released, the currency basket has hovered at 2.53 as of
October 27 (Chart 5.1.8).
Chart 5.1.7.
Chart 5.1.8.
TL and Emerging Market Currencies vs USD*
Currency Basket and The Risk Premium
(22.05.2013=1)
Turkey
Emerging Economies
Selected Emerging Economies
DXY Dollar Index (right axis)
1.40
1.35
Currency Basket (0.5 USD+0.5 euro)
2.8
86
2.7
450
2.6
400
2.5
350
2.4
300
2.3
250
Interim MPC Meeting
1.30
85
1.25
EMBI+Turkey (right axis)
87
500
84
1.20
83
1.15
82
1.05
1.00
0914
0714
0514
0314
0114
1113
0913
0713
0513
0.95
2.2
200
81
2.1
150
80
2
100
79
1.9
50
0112
0312
0512
0712
0912
1112
0113
0313
0513
0713
0913
1113
0114
0314
0514
0714
0914
1.10
* Emerging economies include Brazil, Chile, Colombia, Czech Republic, Hungary, Indonesia, Mexico, Poland, Romania, South Africa, India and Turkey.
Selected emerging economies are Brazil, Indonesia, South Africa and India.
Source: Bloomberg.
Third-quarter developments also had repercussions on the implied exchange rate volatilities of
emerging market currencies causing the implied volatilities of emerging market currencies to increase.
Turkey’s implied exchange rate volatility moved in tandem with other emerging economies and posted
a rise in both short-term and long-term maturities (Chart 5.1.9). Thanks to the recently improved global
risk appetite, the implied volatility of the Turkish lira decreased again by diverging positively from the
exchange rates of peer emerging economies. In addition, the movement observed in the exchange
rate in the beginning of the third quarter was also evident in risk reversal positions, which denote the
gap between the volatilities implied by call and put options. An increase in this gap shows that the
expectations for an appreciation in the Turkish lira outweigh the expectations for depreciation. Yet, it
appears that risk reversal positions have failed to reflect the rise in the exchange rate since September
(Chart 5.1.10.)
52
Inflation Report 2014-IV
Central Bank of the Republic of Turkey
Chart 5.1.9.
Chart 5.1.10.
Implied Volatility of Exchange Rates*
25 Delta Risk Reversal Positions at Various Maturities*
(1-Month-Ahead)
(5-Day Moving Average, Percent)
1-Month
1-Year
7
30
30
Emerging Economies with
Currency Account Deficit
27
27
24
3-Month
USD/TL
2.4
6
2.3
5
2.2
24
Turkey
21
21
1.7
* Emerging economies with current account deficits include Brazil, Chile,
Colombia, Czech Republic, Hungary, Indonesia, Mexico, Poland,
Romania, South Africa and India.
Source: Bloomberg.
0914
1.8
0
0714
1
3
0514
6
3
0314
6
0114
1.9
9
1113
2
9
0913
2
12
0713
3
12
0513
2.1
0313
4
15
0113
18
15
0513
0613
0713
0813
0913
1013
1113
1213
0114
0214
0314
0414
0514
0614
0714
0814
0914
1014
18
*Risk reversal position denotes the difference between implied volatilities of
call and put options with the same delta. An increase indicates higher
expectations for the depreciation of the TL with respect to an
appreciation.
Source: Bloomberg, CBRT.
Monetary Policy
In the third quarter of 2014, global liquidity conditions deteriorated slightly, while the uncertainty
over global monetary policies continued. In this period, while the Fed ended its asset purchases, it also
announced that accommodative policies would be maintained for a while, which caused expected
long-term interest rates to fall. In addition, the ECB cut policy rates against economic slowdown and
risks of deflation and announced the details of the quantitative easing program. This policy divergence
between advanced economies causes a deceleration of economic activity in emerging economies
as well as heightened volatility in capital flows towards these economies. Against this background, the
CBRT maintained its prudent policy stance and delivered measured rate cuts. In this context, the oneweek repo rate was lowered from 8.75 to 8.25 percent while the overnight borrowing rate was reduced
from 8 to 7.5 percent in July. In order to ensure the symmetry of the interest rate corridor, the overnight
lending rate was lowered from 12 to 11.25 percent, while the rate on borrowing facilities for primary
dealers via repo transactions was cut from 11.5 to 10.75 percent.
In order to maintain balanced growth and capital inflows during the upcoming global monetary
policy normalization, the CBRT has changed the foreign exchange deposit rates that apply to banks to
borrow from the CBRT within their limits through the Foreign Exchange Deposit Market. As of October 9,
the rates for one-week maturity borrowings from the CBRT as the last resort facility have been reduced
from 10 percent to 7.5 percent for USD and from 10 percent to 6.5 percent for eur. Moreover,
considering the increase in banks’ balance sheets and the CBRT’s international reserves, it was
announced that banks’ transaction limits at the Foreign Exchange and Banknotes Markets, which are
currently 10.8 billion USD, will be revised at the press meeting for the release of Monetary and Exchange
Rate Policy for 2015. In addition, it was noted that it is crucial to further strengthen the currently solid
structure of the banking sector for financial stability purposes. To this end, the CBRT announced on
October 21 that it will start to remunerate the Turkish lira component of required reserves of financial
institutions
in
November
Inflation Report 2014-IV
2014
to
spur
balanced
growth
and
domestic
savings.
53
Central Bank of the Republic of Turkey
Accordingly, the remuneration rate will be the average CBRT funding cost minus 500 basis points for
banks whose core liability ratios (deposit and equity over loans) are above the sector’s average and
the average CBRT funding cost minus 700 basis points for the remaining banks. Paying interest for core
liabilities is thus considered an incentive.
Since the July Inflation Report, the CBRT continued to fund the market primarily from the oneweek repo rate following the decision made at January’s interim MPC meeting that simplified the
operational framework of monetary policy (Chart 5.1.11). The provision of liquidity primarily by oneweek repo auctions enabled the CBRT average funding rate to near the weekly funding rate.
Meanwhile, in view of the heightened geopolitical tensions and the financial market volatility, the tight
monetary policy stance has been invigorated by a tight liquidity policy since September. Thus, the BIST
overnight repo rates that hovered around the one-week repo auction rate in July and August has lately
settled close to the upper end of the interest rate corridor (Chart 5.1.12).
Chart 5.1.11.
Chart 5.1.12.
CBRT Funding*
CBRT Rates and BIST O/N Repo Rates
(2-Week Moving Average, Billion)
(Percent)
Marginal Funding
O/N Funding
1-Week Repo
1-Month Repo
Reverse Repo at the BIST and IMM
Net OMO
60
15
50
13
13
40
40
11
11
30
30
9
9
20
20
7
7
10
10
5
5
0
0
3
3
-10
1
1
50
0914
0714
0514
0314
0114
1113
0913
0713
0513
0313
0113
-10
15
0811
1011
1211
0212
0412
0612
0812
1012
1212
0213
0413
0613
0813
1013
1213
0214
0414
0614
0814
1014
60
Interest Rate Corridor
CBRT Average Funding Rate
BIST O/N Rates
1-Week Repo Rate
Source: BIST, CBRT.
While the CBRT had implemented measured rate cuts in the third quarter, it has maintained a
tight monetary policy stance by keeping the yield curve nearly flat. The spread between 5-year market
rates and the BIST overnight repo rates is at negative levels as of October (Chart 5.1.13). In addition,
with the recent tightening in the liquidity policy imposed by the CBRT, short-term rates rose substantially,
while the yield curve remained flat (Chart 5.1.14). The CBRT will closely monitor inflation expectations,
pricing behavior and other factors affecting inflation in the upcoming period and maintain its tight
monetary policy stance by keeping the yield curve flat until there is a significant improvement in the
inflation outlook.
54
Inflation Report 2014-IV
Central Bank of the Republic of Turkey
Chart 5.1.13.
Chart 5.1.14.
Market Rates and the CBRT Funding Rate
Yield Curve
(Percent)
(Percent)
5-Year Market Rate-O/N Repo Rate at the BIST
5-Year Market Rate
O/N Repo Rate at the BIST
CBRT Average Funding Rate
October 7-27
9.25
6
9.00
9.00
4
4
8.75
8.75
8.50
8.50
8.25
8.25
8.00
8.00
2
2
0
0
-2
-2
-4
0611
0811
1011
1211
0212
0412
0612
0812
1012
1212
0213
0413
0613
0813
1013
1213
0214
0414
0614
0814
1014
-4
8.88
9.25
6
8.85
8
8.83
9.50
8
8.83
9.50
8.84
9.75
10
8.84
9.75
10
8.75
12
8.53
10.00
12
8.52
10.00
8.47
14
8.45
14
July 25-October 27
Maturity (Year)
Source: BIST, Bloomberg, CBRT.
Source: Bloomberg
In the third quarter of 2014, Turkish lira costs remained higher than foreign currency costs as in
the previous quarter, leaving the use of the ROM advantageous. In fact, banks resorted to both gold
and foreign exchange ROM to a great extent (Charts 5.1.15 and 5.1.16). Banks’ use of ROM stands at
94 percent (56.7/60) for FX and 89 percent (26.8/30) for gold as of the maintenance period starting on
October 24.
Chart 5.1.15.
Chart 5.1.16.
Banks’ Use of ROM for FX*
Banks’ Use of ROM for Gold*
(Percent)
(Percent)
Upper Bound for FX
Use of ROM for FX
Upper Bound for Gold
Use of ROM for Gold
30
30
50
50
25
25
40
40
20
20
30
30
15
15
20
20
10
10
10
10
5
5
0
0
0
0
0911
1111
0112
0312
0512
0712
0812
1012
1212
0213
0413
0613
0813
0913
1113
0114
0314
0514
0714
0814
1014
60
0911
1111
0112
0312
0512
0712
0812
1012
1212
0213
0413
0613
0813
0913
1113
0114
0314
0514
0714
0814
1014
60
* As of the maintenance period.
Source: CBRT.
CBRT reserves remained virtually unchanged compared to the July Inflation Report
(Chart 5.1.17). In the third quarter, considering the global market developments and the operational
procedures for required reserves, the CBRT restricted the type of currency maintained within the ROM
to USD as of the maintenance period which started on August 15. In this period, the marginal decline in
reserves that the banks maintained under the foreign exchange and gold reserve options was
compensated by the increase in foreign currency required reserves. Due to the increased volatility in
exchange rates during the third quarter, the CBRT revised the foreign exchange selling auction amount
from minimum 10 million USD to minimum 40 million USD as of October 29. Furthermore, the CBRT raised
the limits of export rediscount loans and lowered loan costs on October 20 to support the balanced
Inflation Report 2014-IV
55
Central Bank of the Republic of Turkey
growth. These changes are expected to raise exporters’ borrowing from the CBRT, which will also lead
to an increase in the CBRT’s FX reserves in 2015. However, continuing with FX selling auctions will lower
the CBRT’s FX reserves in the upcoming period (Table 5.1.1).
Chart 5.1.17.
Table 5.1.1.
CBRT FX Reserves*
Contribution to FX Reserves*
(Billion USD)
(Billion USD)
Gold Reserves (Precious Metals)
Other Gold Reserves
Gold Reserves (ROM)
FX Reserves (ROM)
FX Reserves (FX Required Reserves)
Other FX Reserves
140
120
FX Sales
(-)
Rediscount
Credits (+)
140
January 2014*
5.8
0.57
120
February 2014
1.0
0.30
March 2014
1.05
0.48
100
100
April 2014
1.02
0.36
80
80
May 2014
0.50
2.06
June 2014
0.42
1.54
July 2014
0.38
1.46
August 2014
0.21
1.38
September 2014
0.31
October 2014
November 2014
December 2014
0.64**
0914
0714
0514
0114
0314
1113
0913
0713
0513
0113
0313
0
1112
0
0912
20
0712
20
0512
40
0312
40
0112
60
1111
60
1.42
1.32***
0.78***
1.32***
* Includes direct FX sales on January 23.
** As of October 27.
*** Provisional.
Source: CBRT
* As of October 24.
Source: CBRT.
Market Rates
Due to the rise in emerging market sovereign risk premiums and the weak and volatile capital
flows in the beginning of the third quarter of 2014, market rates went up slightly in emerging economies.
With the recovery in the global risk appetite in October, the increase in market rates was reversed to
some extent (Charts 5.1.18 and 5.1.19). In this period, Turkey’s market rates firstly rose and then fell.
Turkey’s proximity to regions dealing with geopolitical risks and its commercial ties with the countries in
this region caused market rates to follow a more volatile course in Turkey than in any other emerging
economy. Across countries, Turkey’s 5-year and 6-month market rates are among those which
exhibited the most dramatic increase since the previous reporting period (Charts 5.1.20 and 5.1.21).
Chart 5.1.18.
Chart 5.1.19.
5-Year Market Rates*
6-Month Market Rates*
(Percent)
14
13
(Percent)
Brazil
India
Turkey
South Africa
Indonesia
12
11
South Africa
Indonesia
13
13
12
12
11
11
10
10
9
9
8
8
7
7
6
6
5
5
11
10
Brazil
India
Turkey
14
10
13
12
8
8
7
7
6
6
4
4
5
5
3
3
4
4
2
2
0113
0213
0313
0413
0513
0613
0713
0813
0913
1013
1113
1213
0114
0214
0314
0414
0514
0614
0714
0814
0914
1014
9
0113
0213
0313
0413
0513
0613
0713
0813
0913
1013
1113
1213
0114
0214
0314
0414
0514
0614
0714
0814
0914
1014
9
* As of October 27. 4-year market rates are used for Brazil.
Source: Bloomberg
56
Inflation Report 2014-IV
Central Bank of the Republic of Turkey
Chart 5.1.20.
Chart 5.1.21.
5-Year Market Rates*
6-Month Market Rates*
(Percent)
(Percent)
1.0
1.0
1.0
0.8
0.8
0.8
0.8
0.6
0.6
0.6
0.6
0.4
0.4
0.4
0.4
0.2
0.2
0.2
0.2
0.0
0.0
0.0
0.0
-0.2
-0.2
-0.2
-0.2
-0.4
-0.4
-0.4
-0.4
-0.6
-0.6
-0.6
-0.6
-0.8
-0.8
-0.8
-0.8
-1.0
-1.0
-1.0
-1.0
Brazil
Mexico
Indonesia
Turkey
Peru
Malaysia
Hungary
India
Chile
Colombia
South Africa
South Korea
Tailand
Czech Rep.
Romania
China
Polonya
Turkey
Brazil
Mexico
Malaysia
Colombia
South Africa
Thailand
India
Czech Rep.
Indonesia
China
Peru
Romania
South Korea
Hungary
Chile
Polonya
1.0
* Denotes changes from July 24 to October 29. 4-year rates are used for Brazil.
Source: Bloomberg.
Due to growing geopolitical unrest and financial market volatility since September, the CBRT
resorted to additional liquidity measures besides maintaining a tight monetary policy stance. Thanks to
the recent liquidity measures, the BIST overnight repo rates neared the upper band of the interest rate
corridor. Accordingly, the distribution of expected overnight rates at the BIST Repo and Reverse Repo
Market shifted slightly towards the right compared to July (Chart 5.1.22). Inflation expectations, another
factor that may be influential on market rates, posted a rise compared to July (Chart 5.1.23).
Chart 5.1.22.
Chart 5.1.23.
Expected O/N Rates at the BIST Repo and Reverse
Repo Market*
Inflation Expectations*
July 2014
0.9
October 2014
0.9
0.8
0.8
0.7
0.7
0.6
0.6
0.5
0.5
July 2014
9.5
October 2014
9.5
9.16
9.0
9.0
8.5
8.5
8.0
8.0
8.30
7.54
0.4
0.4
7.5
0.3
0.3
7.0
0.2
0.2
0.1
0.1
7.5
7.27
6.79
6.5
* CBRT Survey of Expectations.
Source: CBRT.
1116
0916
0716
0516
6.0
0316
6.0
0116
13
1115
12
0915
11
0715
10
0515
9
0315
8
0115
7
1114
0
6
6.5
6.73
1014
0
7.0
* End of current month, current year-end, 12-month-ahead and 24-monthahead policy rate expectations derived from the CBRT Survey of
Expectations.
Source: CBRT.
Real rates firstly increased, and then decreased in Turkey during the third quarter of 2014. Even
though two-year inflation expectations remained broadly unchanged, nominal rates have been
influential in the course of two-year real rates in this quarter (Chart 5.1.24). Meanwhile, moving parallel
with Turkey’s risk premium, the benchmark rate rose first in the third quarter, and then declined
(Chart 5.1.24). After the recent decreases, Turkey’s two-year real rate ranked around the middle
among other emerging economies (Chart 5.1.25).
Inflation Report 2014-IV
57
Central Bank of the Republic of Turkey
Chart 5.1.24.
Chart 5.1.25.
2-Year Real Interest Rates for Turkey*
2-Year Real Interest Rates*
(Percent)
(Percent)
Real Interest Rate
Benchmark Rate (right axis)
5
12
4
6
6
5
5
4
4
3
3
2
2
1
1
0
0
-1
-1
-2
-2
11
3
10
2
9
8
1014
0914
0814
0714
0614
0514
0414
0314
0214
0114
1213
1113
1013
6
0913
-1
0813
7
0713
0
Brazil
Colombia
Indonesia
India
China
Peru
Turkey
South Africa
Poland
Hungary
South Korea
Chile
Mexico
Romania
Thailand
Malaysia
Israel
Czech Rep.
Philippines
1
* Calculated as the difference between 2-year market rates obtained from
Bloomberg and the 24-month-ahead inflation expectations derived from the
CBRT Survey of Expectations. As of October 27.
Source: BIST, CBRT.
* Calculated as the difference between 2-year government bond returns of
countries and the 24-month-ahead inflation expectations derived from the
Consensus Forecasts. As of October 29.
Source: Bloomberg, Consensus Forecasts, CBRT.
Loan Rates and Banking Sector Funding Costs
Rates on loans extended to the non-financial sector, which increased notably in early 2014,
decreased gradually the rest of the year due to the loosened domestic and external financing
conditions. The largest fall in consumer loans appeared in housing loans, posting a quarter-on-quarter
decline of 130 basis points (Chart 5.1.26). Commercial loan rates, which are mostly extended in the
short term, continued to fall by about 100 basis points. The drop during the week of the Ramadan
holiday was discontinued with commercial loan rates rebounding the following week (Chart 5.1.27).
According to the Loan Tendency Survey results of the third quarter of the year, banks slightly lowered
coverage ratios for commercial firms, while fees and commissions (non-interest charges) remained
unchanged.
Chart 5.1.26.
Chart 5.1.27.
Consumer Loan Rates
TL Commercial Loan Rates
(Flow, Annualized, 4-Week Moving Average, Percent)
20
5
5
0814
8
0514
8
0214
11
1113
11
0813
14
0513
14
0213
17
1112
17
0512
0814
0514
0214
8
1113
8
0813
10
0513
10
0213
12
1112
12
0812
14
0512
14
0212
16
1111
16
0811
18
0511
18
0211
20
20
0212
22
20
1110
Commercial Loan Rate
Commercial Loan Rate (excl. overdraft accounts)
Housing
1111
Automobile
Personal
22
0812
(Flow, Annualized, Percent)
Source: CBRT.
Rates on deposits, which are the main financing source of the banking sector and are heavily
concentrated on less than 3-month maturity, started the second quarter with a decline, reflecting the
CBRT’s decisions and flattened in August and September. In this period, reflecting the respective
58
Inflation Report 2014-IV
Central Bank of the Republic of Turkey
developments, the spread between the commercial loan rate and the deposit rate went slightly below
the 3.5 percent mark at the end of the third quarter (Chart 5.1.28). The slowing growth across emerging
economies in the third quarter and the uncertainty surrounding global monetary policies caused
capital flows to be weak and volatile in this period, and capital flows to Turkey were no exception due
to increased sovereign risk premiums. Thus, rates on bills and bonds issued by banks posted an increase
in August, and edged down in September (Chart 5.1.29).
Chart 5.1.28.
Chart 5.1.29.
TL Commercial Loan Rate and Deposit Rate
(Flow, Annualized, 4-Week Moving Average, Percent)
Indicators on Banks’ Funding Costs
13
6.5
12
12
6.0
11
11
5.5
10
10
5.0
9
9
4.5
8
8
4.0
7
7
3.5
6
6
3.0
5
5
2.5
4
4
2.0
3
3
15
13
11
9
7
0914
0714
0514
0314
0114
1113
0913
0713
0513
0313
0113
1112
0912
0712
0512
0312
0112
5
Bills and Bonds Rate
Deposit Rate
CBRT Average Funding Rate
13
0811
1011
1211
0212
0412
0612
0812
1012
1212
0213
0413
0613
0813
1013
1213
0214
0414
0614
0814
17
TL Deposit Rate
TL Commercial Loan Rate
TL Commercial Loan Rate - TL Deposit Rate (right axis)
7.0
Source: CBRT.
5.2. Credit Volume and Monetary Indicators
The net credits to the GDP ratio, which is critical to financial stability and an indicator of the
relationship among credit growth, economic activity and aggregate demand, trended further
downwards in the third quarter of 2014 and fell below 12 percent reflecting the slowdown in the credit
growth (Chart 5.2.1). In the next quarter, with the CBRT’s prudent policy stance, the net credits to the
GDP ratio will pare down moderately. Meanwhile, the external net credit use of firms remained close to
historical averages in this period, indicating that firms had easy access to external borrowing
(Chart 5.2.2).
Chart 5.2.1.
Chart 5.2.2.
Domestic Credit Stock and Net Credit Use*
External Credit Stock and Net Credit Use
(Percent)
(Percent)
External Credit Stock/GDP
Domestic Credit Stock/GDP
Net Credit Use/GDP (right axis)
External Net Credit Use/GDP (right axis)
18
10
65
16
9
3.5
60
14
8
3.0
7
2.5
70
55
12
4.0
2.0
6
50
10
45
8
40
6
3
0.0
35
4
2
-0.5
30
2
1
-1.0
25
0
0
1
3
2008
1
3
2009
1
3
2010
1
3
2011
1
3
2012
1
3
2013
1 3**
2014
1.5
5
1.0
4
0.5
-1.5
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3*
2008
2009
2010
2011
2012
2013
2014
* Net credit use is measured as the annual change in nominal credit stock.
** Estimate.
Source: CBRT.
Inflation Report 2014-IV
59
Central Bank of the Republic of Turkey
The annual growth rate of loans extended to the non-financial sector, which has been slowing
due to the CBRT’s tight monetary policy and the BRSA’s measures introduced in the beginning of the
year, recorded a mild decline in the third quarter of 2014. The annualized total loan growth rate
reflecting the loan developments over the past three months neared past years’ averages, albeit
hovering slightly below. This was largely driven by loan rates declining modestly amid falling domestic
funding costs. However, a quarter-on-quarter decrease in consumer confidence and the unchanged
sentiment for the overall economic outlook prevented the loan growth rate from rising. Against these
developments, loans extended to the non-financial sector posted a 16.4 percent year-on-year growth
in exchange rate adjusted terms at the end of the second quarter of 2014 (Chart 5.2.3), while the 13week moving average covering the third quarter recorded a 16.7 percent growth in annualized terms
(Chart 5.2.4). In the final quarter, the improved sentiment for the overall economic outlook amid the
expected rise in households’ and firms’ loan demands implies an increase in loan growth rates, while
banks’ expectations of a pickup in both domestic and external funding costs may restrain this implied
increase.
Chart 5.2.3.
Chart 5.2.4.
Loan Growth
Loan Growth
(Adjusted for Exchange Rate Effect, Annualized, Percent)
16
16
14
14
12
12
10
10
Total
Consumer
20
20
18
15
15
10
10
5
5
0
0
1014
20
0714
25
0414
25
0114
30
1013
30
0713
35
0413
35
0113
40
1012
40
0712
1014
18
0714
20
0414
22
0114
22
1013
24
0713
24
0413
26
0113
26
1012
28
0712
28
0412
30
0112
Commercial
Consumer
0412
Total
30
0112
Commercial
(Adjusted for Exchange Rate Effect,
13-Week Moving Average, Annualized, Percent)
Source: CBRT.
The uptrend of the annualized consumer loan growth that started in May continued into the
third quarter (Chart 5.2.5). The consumer loan growth rate that has been rising moderately despite
seasonality reflects the changes in consumer loan rates that are gradually falling. This upward trend is
more prevalent in housing loans, which has a higher sensitivity to the interest rate. Meanwhile, personal
loans move on par with the past years’ averages. The annualized growth rate of housing loans stood at
17.5 percent at the end of the quarter, while the annualized growth rate of personal loans remained
below the past years’ averages with 18.3 percent in the same period (Chart 5.2.6). According to the
results of the Loan Tendency Survey that covers the third quarter, loan standards were slightly tighter
across all subcategories of consumer loans, while the demand for housing loans rose markedly.
Personal loan standards are not expected to change in the fourth quarter of 2014, while demand for
housing loans and personal loans are envisioned to rise. However, the minor fall in consumer
confidence indices in the third quarter may curb this increase.
60
Inflation Report 2014-IV
Central Bank of the Republic of Turkey
Chart 5.2.5.
Chart 5.2.6.
Consumer Loan Growth
Consumer Loan Growth
(13-Week Moving Average, Annualized, Percent)
2014
(13-Week Moving Average, Annualized, Percent)
2007-2013 Average
25
25
20
20
20
10
10
15
15
0
0
10
10
-10
-10
5
5
-20
-20
0
-30
-30
Dec
Oct
Nov
Sep
Jul
Aug
Jun
Apr
May
Mar
Jan
1014
30
20
0714
30
0414
30
0114
30
1013
40
0713
40
0413
35
Feb
Personal
0113
35
1012
Automobile
50
0
Housing
0712
50
0412
40
0112
40
Source: CBRT.
Having remained slightly lower than seasonal averages in the second quarter, the annualized
growth rate of commercial loans was close to the past years’ averages in the third quarter
(Chart 5.2.7). In terms of currency denomination, the growth rate of TL commercial loans fell, while the
growth rate of FX-denominated commercial loans picked up (Chart 5.2.7). According to the Loan
Tendency Survey results that cover the third quarter, loan standards remained unchanged in terms of
scale, maturity and the type of currency. In this period, the requested collaterals for commercial loans
were slightly down. On the demand side, firms demanded fewer loans from banks. This contraction was
more evident across large-scale firms. Analysis of the factors affecting firms’ loan demand reveals that
the investment demand fell significantly in the third quarter. Inventory build-up and use of working
capital, which were among the main factors boosting demand in the previous period, failed to spur
demand in this period. Banks indicated that they do not expect a change in commercial loan
standards for the final quarter of the year. Meanwhile, firms’ loan demand is likely to increase in the
same period, which applies to the loan demand from both SMEs and large-scale firms. Yet, banks
expect the demand for short-term loans to remain unchanged.
Chart 5.2.7.
Chart 5.2.8.
Commercial Loan Growth
TL and FX Commercial Loan Growth
(Adjusted for Exchange Rate Effect, 13-Week Moving
Average, Annualized, Percent)
2007-2013 Average
(13-Week Moving Average, Annualized, Percent)
TL Commercial Loans
2014
40
40
45
FX Commercial Loans (including foreign branches)
45
35
35
40
40
35
35
30
30
30
30
25
25
25
25
20
20
20
20
15
15
10
10
15
15
1014
0714
0414
0114
1013
0713
0413
0113
1012
0712
0412
0112
Dec
Nov
Oct
-5
Sep
-5
Aug
0
Jul
0
0
Jun
0
May
5
Apr
5
5
Mar
5
Feb
10
Jan
10
Source: CBRT.
Inflation Report 2014-IV
61
Central Bank of the Republic of Turkey
In the third quarter of 2014, growth rate of loans, especially consumer loans, continued to
decline in annualized terms. Macroprudential measures led to a faster growth in commercial loans
compared to consumer loans, supporting the balancing of the economy. This is consistent with an
outlook in which domestic demand has little support for economic activity and net exports make a
greater contribution. The expected rise in both domestic and external funding costs and the course of
consumer confidence are estimated to bring loan growth rates down, while the impending increase in
the demand for consumer and commercial loans and the likely economic recovery may drive loan
growth rates higher in the upcoming quarter. The loan growth that has been slowing due to policies
adopted by the CBRT and the BRSA is expected to gradually near the deposit growth rate (Chart
5.2.9). The decline in the difference between loan and deposit growth is a factor that will enhance the
resilience of the banking sector against possible financial fluctuations by also reducing the banking
sector’s need for external financing.
Chart 5.2.9
Growth of Deposits and Loans*
(Annual Change/GDP)
18
Deposits
Loans
18
16
14
14
12
12
10
10
8
8
6
6
4
4
2
2
0
0
0109
0409
0709
1009
0110
0410
0710
1010
0111
0411
0711
1011
0112
0412
0712
1012
0113
0413
0713
1013
0114
0414
0714
16
* Including participation banks, excluding interbank deposit and not adjusted for exchange
rate effect.
Source: TurkStat, CBRT.
Monetary Indicators
The uptrend in credits extended to the private sector continued to determine the annual growth
of M3, the broad measure of money supply, in the third quarter of 2014. The rate of increase in the
Private Sector Claims mostly including the credits extended by banks to non-financial private
individuals and institutions continued to fall, constituting the main factor of the decline in the growth of
the M3.
In the third quarter, Public Sector Claims continued to contribute positively to the M3 growth for
the second quarter in a row. The negative contribution of net external assets posted a quarter-onquarter decrease. Meanwhile, the negative contribution of the item Other, which displayed a relatively
steady course in line with bank profitability, is still a non-deposit funding source for the banking sector,
yet recorded a slight fall compared to the end of the second quarter (Chart 5.2.10).
62
Inflation Report 2014-IV
Central Bank of the Republic of Turkey
Chart 5.2.10.
Chart 5.2.11.
Balance Sheet Decomposition of M3
Currency in Circulation and Current Consumption
Spending*
(Contributions to Annual M3 Growth)
4. Other
3. Private Sector Claims
2. Public Sector Claims
1. Net External Assets
1+2+3-4= M3 (annual percent change)
40
(Seasonally Adjusted)
Current Consumption Spending (annual percent change)
Currency in Circulation (annual percent change)
Currency in Circulation (right axis)
Percent
40
30
30
20
20
10
10
0
0
-10
-10
-20
-20
BillionTL
35
95
30
85
25
75
20
65
15
55
10
45
5
35
0507
0907
0108
0508
0908
0109
0509
0909
0110
0510
0910
0111
0511
0911
0112
0512
0912
0113
0513
0913
0114
0514
0
Source: CBRT.
25
123412341234123412341234123
2008
2009
2010
2011
2012
2013 2014
* Consumption spending includes private and public consumption
excluding furniture, household appliances, and transport and
communication services at current prices.
Source: TurkStat, CBRT.
The annual growth of seasonally adjusted currency in circulation recorded a quarter-on-quarter
increase in the third quarter of 2014 (Chart 5.2.11). This is in line with the expected third-quarter growth
in private demand. Yet, the expected moderation in domestic demand and the slowdown in
consumer loan growth due to the adopted macroprudential measures, the tight monetary policy
stance and the weak capital flows fuel expectations of a more modest increase in the annual growth
of the currency in circulation.
Inflation Report 2014-IV
63
Central Bank of the Republic of Turkey
Box
5.1
Non-Core Liabilities
Total loans extended to the private sector increased at a relatively faster rate in emerging economies than
advanced economies during the pre-crisis period. Although the global financial crisis put a cap on loan
growth in emerging economies for a while, loans have picked up, particularly in recent years. This increase
is largely attributed to the ample pre-crisis liquidity that became even more abundant due to post-crisis
quantitative easing policies adopted by advanced economies.
Although credit expansion has the ability to absorb postponed loan demand, and drive investments and
thus income higher, a fast credit expansion is accompanied by potential risks. One of these risks is the fact
that loans can dampen a current account balance through import demand. Another risk, as we learned
from the last global crisis, is that the bubbles occurring in asset prices due to a fast growing purchasing
power might lead to systemic crises through the financial accelerator effect.1 The macroprudential policies
that are frequently referred to by central banks in recent years thus emphasize the importance of taking
into consideration the heightened fragility in periods of easing when expectations are benign. To this end,
central banks have adopted new policies that combine the conventional objective of price stability with
financial stability.
The
question of how to design a macroprudential framework is relevant in globalizing and deepening
financial markets. Accordingly, every economy has to take individual action with respect to their own
financial conditions; however, there are also attempts to construct an international framework.2 Borio (2010)
states that studies regarding macroprudential policies involve two main approaches based on the
distribution of the size of the risk. The first approach studies the development of the systemic risk in financial
markets over time and emphasizes that it is crucial to implement countercyclical policies in times of high
risk. The other approach deals with how the risk is distributed in a certain time within the financial network
determined by multiple financial structures and with the correlation among the risks of financial institutions,
and focuses on measures that prevent the impact of a likely financial market risk from surging due to
negative externalities.
As mentioned above, there are many risk indicators that address the development of financial market risks
both over time and from a cross-sectional point of view. Some of these indicators are early warning systems
dependent on market data. Another approach that has gained wide support in the literature in recent
years aims to move the popular pre-crisis macroprudential perspective based on the balance sheets of
financial institutions to a macroprudential framework that takes into account the increasingly more
sophisticated structure of financial markets. This approach puts more emphasis on stress tests, financial
network models or econometric models. Against this background, this box provides information on noncore liabilities and related studies in Turkey for non-core liabilities, which have been recently suggested as a
systemic risk indicator in the literature.
Schularick and Taylor (2014) show the correlation between credit expansion and financial crises for 14 countries during a time span of more than
130 years.
2 The BIS publishes brief reports at certain intervals on how the international BASEL regulatory policy framework is implemented in respective
countries. For an up-to-date report, see BIS (2014).
1
64
Inflation Report 2014-IV
Central Bank of the Republic of Turkey
Shin (2010) suggests that the reliability of bank funding should be analyzed by decomposing the liability
side of the balance sheets of financial institutions. Defining core and non-core liabilities is useful in
understanding through what channels the credit expansion is financed during periods of economic
recovery. To meet the loan demand in times of economic recovery, instead of reliable funding sources
such as deposits, banks may opt for external debt or relatively less reliable and volatile funds such as repo.
In terms of a more sophisticated financial network structure and an increased interdependence between
financial institutions, such a lending behavior imposes more fragility on the sector against systemic risks. In
this regard, Akdoğan and Yıldırım (2014) present a classification of core and non-core balance sheet
liabilities for Turkey3,4 (Table 1).
Table 1. Classification of Core and Non-Core Liabilities
Core Liabilities
Liquid
Partly Liquid
Non-Liquid
Household
Demand deposits
Short-term deposits (up to 1month)
Intermediate
Non-Financial Corporations
Demand deposits
Short-term deposits
(up to 3-month)
Non-Core Liabilities
Financial Institutions
Demand deposits
Funds from repo
Short-term payables to banks
Medium-term deposits
(1-month to 1-year)
Long-term deposits
(1-year and more)
Medium and long-term
deposits
Medium and long-term deposits
Medium and long-term payables to banks
Securities Issued
Other borrowings from banks
Source: Akdoğan and Yıldırım (2014).
As shown in Table 1, deposits of domestic households, which are relatively reliable and costless, constitute
core funding sources that move in tandem with the income level of households. However, the increase in
loan demand may be higher than the increase in deposits in times of economic recovery. In these periods,
instead of reliable funds such as deposits, banks may prefer to provide funding from other financial
institutions, particularly external ones. Therefore, non-core liabilities include debt items, such as payables to
banks, repo, etc., borrowed from other financial institutions. These liabilities are funding resources that are
both shorter term and more volatile than core liabilities.
Regarding the components of non-core liabilities for Turkey, the payables to banks item accounts for a
larger share than repo transactions (Chart 1). In terms of the type of borrowing, the payables to banks item
consists more of borrowing from external banks, while repo transactions are mostly composed of Turkish lira
transactions. Chart 2 shows the ratio of non-core liabilities to total external funds.5 Having hovered between
17-22 percent during 2005-2010, this ratio increased as high as 25 percent afterwards.
Akdoğan and Yıldırım (2014) also examine the course of non-core funds by a comparative analysis on Turkey, US and Korea. The authors find that
the main driver of the rise in non-core funds is repo for the US and external borrowing for Turkey and Korea. The study simulates the Turkish banking
system’s interdependence through a financial networking system, and conducts a liquidity stress test to examine the resilience of the system
against liquidity shocks. Having risen in recent years, non-core liabilities account for about 20 percent of total external funds, which is not an
alarming level for the Turkish banking system as of the period in question; yet it is still advised to use non-core liabilities, which move in tandem with
cyclical volatilities, as a systemic risk indicator.
4 Kılınç et al. (2013) point to a close relationship between non-core liabilities and credit growth in Turkey. Moreover, Özen et al. (2013) examine the
course of non-core liabilities over sub-periods of 1995-2000 and 2004-2012 based on the thesis that non-core liabilities and portfolio flows may
diverge when financial markets are calm or turbulent. The results of the study show that non-core liabilities responded more aggressively to the
financial tightening caused by external developments in the 2004-2012 period than in the former period. Binici and Köksal (2013) indicate that noncore liabilities are among the main determinants of cycles of leverage in the Turkish banking system.
5 Shin (2010) suggests that the ratio of non-core liabilities to M2 should also be taken into account for considering the banking sector’s liabilities
towards depositors.
3
Inflation Report 2014-IV
65
Central Bank of the Republic of Turkey
Chart 1. Sources and Components of Non-Core Funding
Chart 2. Ratio of Non-Core Funding to Total External Funding and
M2 (3-Month Moving Average, Percent)
BillionTL
0314
0913
0313
0912
0312
0911
0311
0910
0310
0909
0309
0908
0308
0907
0307
0906
0306
0714
0114
0713
0113
0712
0112
0711
0111
0710
0110
0709
0109
0708
15
0108
0
0707
20
0107
50
0706
25
0106
100
0705
30
0105
150
0905
35
Milya
200
Non-Core Funding/M2
40
0305
250
Non-Core Funding/Total External Funding
45
Total Payables to Banks
Total Funding via Repo Transactions
Total TL Payables to Banks
Total FX Payables to Banks
Total TL Funding via Repo Transactions
Total FX Funding via Repo Transactions
300
Source: BRSA, CBRT.
Monitoring
the ratio of core liabilities to total external funds as a systemic risk indicator and establishing
policies to bring such liabilities down to reasonable levels will serve to enhance the resilience of the financial
system against liquidity shocks, provide a balanced growth and strengthen domestic savings. Thus, the CBRT
created an incentive in October that remunerates financial institutions whose core liability ratios are higher
than the sector’s average at a higher rate unless they worsen their own situation. 6 The CBRT will closely
monitor both this ratio and the development of the loans to deposits ratio in the upcoming period and adopt
policy tools that will reduce the sensitivity of the financial system to cyclical fluctuations.
REFERENCES
Akdoğan, K. and B.D. Yildirim, 2014, Non-core Liabilities as an Indicator of Systemic Risk and a Liquidity Stress
Test Application on Turkish Banking System, İktisat İşletme ve Finans, 29(338): 39-66.
BIS, 2014, Progress report on implementation of the Basel regulatory framework, Basel Committee on Banking
Supervision.
Binici, M. and B. Köksal, 2012, Is the Leverage of Turkish Banks Procyclical ? Central Bank Review, 12(2): 11-24.
Borio, C., 2010, Implementing a macroprudential framework: Blending boldness and realism, Keynote address
for the BIS-HKMA research conference on Financial Stability: Towards a Macroprudential Approach,
Honk Kong SAR, 5-6 July 2010.
CBRT, 2014, Press Release on Support for Core Liabilities dated October 21, 2014.
Kilinc, Z., H.G. Karasoy and E. Yücel, 2013, Non-Core Bank Liabilities and Credit Growth: Evidence From an
Emerging Economy, International Finance Review, vol. 14, Global Banking, Financial Markets and
Crises, edited by Bang Nam Jeon and Maria Pia Olivero, Chapter 3: 71-90.
Özen, E., C. Şahin and İ. Ünalmis, 2013, External Financial Stress and External Financing Vulnerability in Turkey:
Some Policy Implications for Financial Stability. Central Bank Review, 13(Special I): 65-74.
Schularick, M. and A.M. Taylor, 2009, Credit booms gone bust: monetary policy, leverage cycles and financial
crises, 1870–2008, NBER Working Paper No. 15512.
Shin, H.S. and K. Shin, 2010, Procylicality and Monetary Aggregates, NBER Working Paper No.16836.
6
For details, see CBRT (2014).
66
Inflation Report 2014-IV
Central Bank of the Republic of Turkey
Box
5.2
International Capital Flows and Domestic Loan Growth
International
capital flows have a potential to positively affect the economy in terms of financing
investments, especially in countries with low savings rates. Yet, they can also lead to the accumulation of
macrofinancial risks in emerging economies as experienced in the post-crisis period. Therefore, the
relationship between international capital flows and macrofinancial stability remains a key issue to be
explored. Specifically, benefiting from capital flows while also averting from their negative aspects is a major
challenge faced by policymakers.
This
box examines the empirical relationship between international capital flows and domestic loan growth
by using a panel data set involving 101 emerging and advanced economies.7 Empirical results show a
statistically significant and positive interaction between these two variables. By GDP per capita breakdown,
the loan growth-capital flows correlation is much stronger in upper-middle income countries, including Turkey,
than in lower-middle (LMIC) and higher income countries (HIC).
Chart 1. International Capital Flows and Domestic Loan Growth*
(Percent)
Norway
Japan
-40
3
2.5
2
1.5
1
0.5
0
-0.5 0
-1
-1.5
-2
-2.5
-20
USA
3
15
2
10
1
5
0
0
20
-10
-5
0
5
10
15
-20
0
-5
-1
-10
-2
-15
-3
Turkey
South Africa
Poland
20
3
15
2
6
4
2
10
1
5
-10
-5 0
10
20
-40
-20
-10
-1
0
-2 0
-10
0
0
-20
0
20
40
-6
-8
-3
-20
-10
Thailand
India
Colombia
8
6
3
10
2
5
4
1
2
0
-10
-2 0
0
-40
0
10
20
-2
10
-4
-2
-15
-20
20
0
2
4
6
-20
0
20
40
-5
-1
-4
-6
-2
-10
-8
-3
-15
* Horizontal axis denotes loans/GDP growth, while vertical axis denotes net capital flows/GDP growth.
7For
further details, see Arslan and Taşkın (2014).
Inflation Report 2014-IV
67
Central Bank of the Republic of Turkey
The
data used in the analysis are obtained from International Financial Statistics (IFS), World Development
Indicators (WDI), Chinn and Ito (2006), Darvas (2012) and Broner et al. (2013), covering the period from 1970
to 2009. The analysis is based on macroeconomic variables such as domestic loan growth, net capital flows,
GDP, GDP per capita, real exchange rate, inflation and the external vulnerability index.8
Regression Analysis
Panel data regressions are used to examine the empirical relationship between international capital flows
and domestic loan growth. The baseline regression is estimated using fixed effects model and alternatively
by the Arellano-Bond model. The regression analysis is based on the following equation:
CRit – CRit-1 = α + (NCFit – NCFit-1)β + (Xit - Xit-1)δ + Dyrθ + Dcrλ + μi + εit
(1)
Here, CR and NCF represent the natural logarithm of domestic loans and the international capital flows to
GDP ratio, respectively. X stands for consumer price index, real effective exchange rate, external
vulnerability, real GDP and real GDP per capita. Dyr and Dcr are the year dummy and crisis dummy
variables, respectively.
Table 1. Regression Results
(1)
(2)
(3)
(4)
(5)
(6)
Net Capital Flows/
Variables
0.43***
0.39***
0.37***
0.35***
0.37***
0.37***
Change in GDP
[0.084]
[0.081]
[0.068]
[0.064]
[0.062]
[0.062]
0.62
0.99***
1.16***
1.14***
1.16***
1.18***
Real GDP
[0.514]
[0.176]
[0.188]
[0.188]
[0.178]
[0.180]
Inflation
-0.01**
-0.01**
-0.00
-0.00
-0.00
[0.004]
[0.004]
[0.003]
[0.003]
[0.003]
-0.04*
-0.03*
-0.05**
-0.04
[0.021]
[0.020]
[0.024]
[0.024]
-0.12
-0.10
-0.11
[0.073]
Percentage Change in Real GDP
Crisis dummy
Percentage Change in Real Exchange Rate
Exchange Rate
[0.077]
[0.077]
Change in External
0.01
0.01
Vulnerability Index
[0.008]
[0.008]
Percentage Change in
GDP per capita
Year dummies
0.46
[0.446]
+
+
+
+
+
+
Number of observations
2.187
2.268
2.436
2.436
2.466
2.466
R2
0.16
0.15
0.14
0.13
0.13
0.13
98
100
101
101
101
101
Number of countries
Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1
The
main equation is estimated by fixed effects model using the percentage change values of variables
and also including dummy variables. The estimated coefficients are displayed in Table 1. The first row
displays the estimated value of the coefficient β that shows the correlation between net international
capital flows and loan growth. Each column in the table displays how the results change when explanatory
variables are omitted. The correlation between these two variables is positive (0.43) and statistically
significant. This result remained unchanged when alternative control variables are used (Table 1, columns 16). This estimation also reveals that there are statistically significant coefficients between other explanatory
variables and loan growth (Table 1, rows 2-7).
8 For further details, see Arslan and Taşkın (2014).
68
Inflation Report 2014-IV
Central Bank of the Republic of Turkey
In
order to analyze the correlation between international capital flows and domestic loan growth by
groups of GDP per capita, interaction variables were added to equation (1). Dummy variables were
created for the income group (the higher income group and the lower-middle income group) and the
interaction between these variables and other explanatory variables was added to the equation. The
coefficient estimations of the interaction variables are shown in Table 2. The coefficients of both interaction
variables were negative and statistically significant.
Thus,
the correlation between net capital flows and loan growth is relatively weaker in lower-middle
income countries than in high-middle income countries including Turkey (Table 2, rows 2-3).
Table 2. The Effect of GDP Per Capita
(1)
Percentage
Change in
Loans
0.71***
[0.212]
-0.38
[0.234]
-0.39*
[0.230]
-
Variables
Net Capital Flows/Change in GDP
(2)
Percentage
Change in
Loans
-
Net Capital Flows/ Change in GDP*HIC
-
Net Capital Flows/ Change in GDP*LMIC
-
Net Capital Flows/Trend of Change in GDP
0.77***
[0.218]
-0.52**
[0.245]
-0.45*
[0.236]
-
Net Capital Flows/ Trend of Change in GDP*HIC
-
Net Capital Flows/ Trend of Change in GDP*LMIC
2,176
0.17
97
Number of observations
R2
Number of countries
2.179
0.17
97
Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1
In
sum, this box has examined the relationship between international capital flows and domestic loan
growth. Empirical findings have shown that there is a positive and statistically significant relation between
these two variables. By GDP per capita breakdown, the correlation between domestic loan growth and
capital flows is stronger in upper-middle income economies including Turkey.
REFERENCES
Arellano, J. and S. Bond, 1991, Some Tests of Specification for Panel Data: Monte Carlo Evidence and an
Application to Employment Equation, Review of Economic Studies, 58(2): 277-297.
Arslan, Y. and T. Taşkın, 2014, International Evidence on the Interaction Between Cross-Border Capital Flows
and Domestic Credit Growth, İktisat, İşletme ve Finans, 29(341): 37-56.
Broner, F., T. Didier, A. Erce and S. Schmukler, 2013, Gross Capital Flows: Dynamics and Crises, Journal of
Monetary Economics, 60(1): 113-133.
Chinn, M. and H. Ito, 2006, What Matters for Financial Development? Capital Controls, Institutions, and
Interactions, Journal of Development Economics, 81(1): 163-192.
Darvas, Z., 2012, Real Effective Exchange Rates for 178 Countries: a New Database, Bruegel Working Paper
No. 716.
Inflation Report 2014-IV
69
Central Bank of the Republic of Turkey
Box
5.3
Capital Flows and Loan Growth: The Impact of Macroprudential Measures
Many emerging economies were challenged by problems arising from short-term and volatile capital flows
due to the quantitative easing policies adopted by advanced economies after the global financial crisis.
Capital flows can support economic welfare by allowing converging economies to finance efficient
projects, but they can also lead to financial risks at the same time. Therefore, many emerging economies
have adopted policy measures against capital flow fluctuations in recent years.
This
box analyzes how the elasticity of loan growth to cross-border banking flows has changed after
adopting macroprudential measures against capital flows.9 To this end, a panel data set that covers the
2004Q1-2013Q2 period for 33 advanced and emerging economies is used.10 The findings show that crossborder banking flows is a significant explanatory variable for loan growth, which has become less elastic to
capital flows due to the adoption of macroprudential measures by many emerging economies.
Chart 1. Real Loan Growth and Cross-Border Banking Flows /GDP*
Emerging Economies
Real Loan Growth (Percent)
Real Loan Growth (Percent)
All Countries
Cross-Border Banking Flows/GDP (Percent)
Cross-Border Banking Flows/GDP (Percent)
* The straight line and the dotted line denote simple regression relation during 2004Q1-2010Q3 and 2010Q4-2013Q2, respectively.
Chart 1 shows the real loan growth and the banking-sector capital flows used in the analysis. There is a
significantly positive correlation between the two variables for the 2004Q1-2010Q3 period. This correlation is
stronger for emerging economies. Moreover, this relationship has weakened across all countries and
emerging economies after the end of 2010. Meanwhile, a regression analysis is used to measure how the
correlation between loan growth and capital flows across economies implementing macroprudential
policies has differed from other economies in the post-2010 period.
For further details, see Aysan et al. (2014).
18 emerging economies include Brazil, Bulgaria, Chile, Colombia, Croatia, Czech Republic, Hungary, Indonesia, South Korea, Malaysia, Mexico,
Philippines, Poland, Romania, Russia, South Africa, Thailand and Turkey. 15 advanced economies are Austria, Belgium, Canada, Denmark, Finland,
France, Germany, Iceland, Ireland, Italy, Japan, Netherlands, Portugal, Spain and Sweden.
9
10
70
Inflation Report 2014-IV
Central Bank of the Republic of Turkey
In
the regression analysis, the dependent variable is the quarterly real loan growth (ΔCredits), and the
explanatory variables are the quarterly change of the cross-border banking flows to GDP ratio
(ΔCapitalFlowsit), the quarterly change of the public debt to GDP ratio (ΔPublicDebtit), the quarterly
percentage change of the real exchange rate (ΔRealExchangeit), and interest and inflation rates. The
country dummy variable has a value of 1 for each country selected. Accordingly, the following equation is
calculated using the system GMM method, and overall results and results pertaining to Turkey are given in
Table 1.
ΔCredits𝑖,𝑡 =
𝑎 + ∑2𝑘=1 𝜌𝑘 ΔCredit𝑠𝑖,𝑡−𝑘 + ∑1𝑠=0 (𝛾1𝑠 Δ𝐶𝑎𝑝𝑖𝑡𝑎𝑙𝐹𝑙𝑜𝑤𝑠𝑖,𝑡−𝑠 + 𝛾2𝑠 Δ𝐶𝑎𝑝𝑖𝑡𝑎𝑙𝐹𝑙𝑜𝑤𝑠𝑖,𝑡−𝑠 ∗ Dummy[Macroprudential] +
𝛾3𝑠 Δ𝐶𝑎𝑝𝑖𝑡𝑎𝑙𝐹𝑙𝑜𝑤𝑠𝑖,𝑡−𝑠 ∗ Dummy[Country] + 𝛾4𝑠 Δ𝐶𝑎𝑝𝑖𝑡𝑎𝑙𝐹𝑙𝑜𝑤𝑠𝑖,𝑡−𝑠 ∗ Dummy[Macroprudential] ∗ Dummy[Country]) +
∆PublicDebt/GDP𝑖,𝑡−1 + ∆RealExchange𝑖,𝑡−1 +Inflation𝑖,𝑡−1 + 𝐼𝑛𝑡𝑒𝑟𝑒𝑠𝑡𝑖,𝑡−1 + 𝜇𝑖 + 𝑣𝑖,𝑡
Macroprudential
measures are implemented in 2010Q4 and onwards. The Fed announced its first
quantitative easing package in November 2008, leading to a subsequent abundance of global liquidity,
while emerging economies attracted strong capital inflows and faced currency appreciations. The Fed
announced its second quantitative easing package in November 2010. As of end-2010, many emerging
economies have re-adjusted their policy frameworks to focus on financial stability and adopted new policy
measures. For example, Brazil has adopted tax measures against capital flows; Colombia has imposed
restrictions on off-balance sheet derivatives of banks; Indonesia has raised required reserves and limited
short-term foreign currency borrowing by banks; Korea has restricted the non-core liabilities of banks, and
Thailand has adopted macroprudential measures such as imposing restrictions on the debt-to-value ratio for
the housing market (IMF, 2011).
Table 1. Regression Results
Dependent Variable: Real Loan Growth
Explanatory Variables
Credits(t-1)
(1)
(2)
0.370***
0.351***
[0.066]
[0.072]
-0.255***
-0.251***
[0.036]
[0.038]
0.030***
0.042***
[0.012]
Explanatory Variables (Cont.)
(2)
-
-0.043***
-
[0.014]
-
0.846***
-
[0.086]
ΔCapital Flows (t)*
-
-0.110
[0.015]
Dummy[Macroprudential]*Dummy[Turkey]
-
[0.079]
0.028**
0.037**
ΔCapitalFlows (t-1)*Dummy[Macroprudential]
-
-0.032**
[0.013]
[0.016]
-
[0.014]
-0.220**
-0.236***
-
-0.166**
[0.088]
[0.078]
-
[0.078]
-0.216**
-0.181**
ΔCapitalFlows (t-1)*
-
-0.625***
[0.081]
[0.082]
Dummy[Macroprudential]*Dummy[Turkey]
-
[0.164]
0.059
0.040
Dummy[Macroprudential]
0.535
0.270
[0.057]
[0.061]
[0.525]
[0.457]
1.011***
1.102***
[0.247]
[0.262]
Number of observation
1.229
1.229
Number of countries
33
33
Credits(t-2)
ΔCapitalFlows (t)
ΔCapitalFlows (t-1)
ΔPublicDebt/GDP (t-1)
ΔRealExchange (t-1)
Interest (t-1)
Inflation (t-1)
Δ Capital Flows (t)*Dummy[Macroprudential]
(1)
ΔCapitalFlows (t)*Dummy[Turkey]
ΔCapitalFlows (t-1)*Dummy[Turkey]
Dummy[Global Financial Crisis]
3.142***
[0.966]
3.312***
[0.957]
Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1
Inflation Report 2014-IV
71
Central Bank of the Republic of Turkey
Equation 1 in Table 1 shows the correlation between real loan growth and other macro variables for all
economies and all periods. Accordingly, both the current period capital flows and the lagged value of
capital flows are substantially influential on real loan growth. Having performed a similar analysis by using
annual data and a broader selection of countries and capital flows, Box 5.2 shows that this relationship is
significant, concurring with the findings of Arslan and Taşkın (2014). In equation 2, dummies for the
macroprudential period and Turkey were added to the overall equation. Accordingly, in the period when
macroprudential policies were applied, the elasticity of real loan growth to capital flows is mostly lower for
all economies (coefficients of ΔCapitalFlows*Dummy[Macroprudential] variables). Moreover, this elasticity is
higher
for
Turkey
over
the
whole
period
compared
to
other
economies
(coefficients
of
ΔCapitalFlows*Dummy[Turkey] variables). Thus, across all countries, Turkey stands out with a higher loans-tocapital flows elasticity. Yet, in the period of 2010Q4 and onwards, this elasticity becomes markedly smaller in
Turkey than in other economies (coefficients of ΔCapitalFlows*Dummy[Macroprudential] *Dummy[Turkey]
variables). Even when only emerging economies are used in the regressions, this result is still valid, implying
that Turkey has experienced a lower real loan growth-to-capital flows elasticity during 2010Q4 and onwards
compared to other emerging economies. When the same analysis is repeated for certain emerging
economies, the loans-to-capital flows elasticity is higher in Colombia, Indonesia and Korea over the entire
period compared to other economies but lessens during the implementation of macroprudential measures
as of 2010Q4 and onwards, similar to Turkey.11
In sum, this box concludes that there is a significant correlation between domestic real loan growth and
cross-border banking flows for a broad selection of countries. Many emerging economies have enhanced
their policy frameworks as of 2010 to limit the financial risks arising from volatile and short-term capital flows
and adopted macroprudential measures. The analysis shows that the periods when these measures were
applied are marked by lower loans-to-capital flows elasticity in these countries.
REFERENCES
Arslan, Y. and T. Taşkın, 2014, International Evidence on the Interaction Between Cross-Border Capital Flows
and Domestic Credit Growth, İktisat, İşletme ve Finans, 29(341): 37-56.
Aysan, A.F., S. Fendoğlu, M. Kılınç and S. Yıldız, 2014, Credit Cycles and Capital Flows: Effectiveness of
Macroprudential Policy Framework in Emerging Countries, unpublished manuscript.
IMF, 2011, Recent Experiences in Managing Capital Inflows: Cross-Cutting Themes and Possible Policy
Framework, Public Information Notice No. 11/42.
Similar results were obtained using alternative estimation methods (e.g. fixed effects or differenced GMM). Moreover, the results were similar
even when the macroprudential period was set to a quarter earlier or later than 2010Q4. For details, see Aysan et al. (2014).
11
72
Inflation Report 2014-IV