سیاست های پولی و کارت های اعتباری : شواهدی از یک اقتصاد کوچک باز

This paper uses a unique monthly data set that covers the overall credit card usage in a small-open economy, Turkey, to investigate a possible credit channel of monetary policy transmission through credit cards. A reduced-form vector autoregression analysis is employed, where the forecast error variance decompositions are calculated for three-year windows over the period 2002–2009. It is shown that, during the recent financial crisis that has started in 2007, the monetary policy of Turkey has shifted toward focusing on output volatility and interest rate smoothing through setting short-term interest rates, while the inflation rate has been mostly affected by exchange rate movements and inflation inertia. Credit card usage has an increasing effect on inflation rates through time, requiring more policy emphasis on the credit channel through credit cards. When the effects of the credit view and the money view are compared, the former seems to be more effective on the real side of the economy, independent of the level of inflation.

There is a well-known debate between the money view and the credit view of monetary policy. The former, by assuming that all nonmoney assets are perfect substitutes, emphasizes the money supply affecting aggregate demand through short-term interest rates. The latter goes one step further and states by assuming that bank loans are imperfect substitutes to other nonmoney assets and that the money supply affects aggregate demand not only through short-term interest rates but also through new bank loans. 1 Within this context, the contemporary payment methods, such as credit cards, come into the picture, emphasizing more the role of credit view—so called the credit channel of monetary policy transmission. 2 In particular, in the credit channel of monetary policy transmission, a contractionary monetary policy can force constrained financial institutions (constrained on both sides of their balance sheet) to restrict lending through credit cards, independent of the demand for loans.3 For borrowers dependent on financial institutions through credit cards, the contractionary monetary policy restricts their source of credit and increases the costs of seeking alternative sources.4 Therefore, the most compelling literature on lending effects focuses on the relationship between small banks and small borrowers. Kashyap and Stein, 1995, Kashyap and Stein, 2000, Gertler and Gilchrest, 1993 and Gertler and Gilchrest, 1994 provide a combination of empirical evidence showing that small commercial banks do contract lending after a negative policy shock, and small firms are affected by that contraction.5 The “small” commercial bank assumption is important, since it is assumed that the small borrower relies on a special relationship with the small bank for its credit and finds it difficult to get credit from larger banks or from nonbank alternatives; this is the exact case in the market for credit cards. Within the big picture of credit markets, this paper investigates possible interactions between credit card usage and the monetary policy in a small-open economy, Turkey, which has experienced an increasing credit card usage, especially after 2002. In particular, the volume of transaction with credit cards (i.e., credit card usage) has increased from 24 billion Turkish liras (16 billion euros) in 2002 to 200 billion Turkish liras (92 billion euros) in 2009. During the same period, the real volume of credit card transactions (defined as credit card usage in Turkish liras over the Turkish consumer price index) has increased by 465%, while the number of transactions has increased from 630 million to 1.83 billion; these correspond to an increase in the GDP share of credit card usage from 6% to 20%. On the consumer side, the number of credit cards in Turkey has increased from 14 million to 45 million during 2002–2009; this corresponds to a credit-card-per-capita measure of 0.6 in 2010. On the firm side, the number of POS machines has increased from 0.3 million to 1.7 million during 2002–2009. More importantly, the share of credit card usage within the overall consumer credit market is, on the average, around 20% during 2002–2009. From a policy maker's point of view, these improvements require an increasing focus on the credit channel of monetary policy transmission through credit cards.6 Although underlying factors that influence individual credit card usage may mostly be at the micro level as discussed above, the aggregate credit card usage can be affected by the following macroeconomic variables: (i) prices as one of the most significant factors through the quantity theory of money; (ii) interest rates representing the opportunity cost of credit card usage, especially when credit card debt is not paid on time; (iii) exchange rates as one of the main determinants of prices through international trade; and (iv) production level as a measure for overall health of the economy. For sure, these are interacting factors/variables (e.g., credit card usage may affect prices through the high demand of individuals financed by credit cards, monetary authority may use short-term interest rates for the sake of price stability, which may further credit card usage through both interest rates and prices with opposite signs, or credit card usage may lead to more expensive expenditures such as internationally traded products, which, in turn, will affect exchange rates, prices, interest rates, and output); hence, a reduced-from vector autoregression (VAR) framework is employed to analyze the credit channel of monetary transmission mechanism through credit card usage. Considering the sample period of 2002–2009, the VAR framework will not only allow obtaining impulse response functions and variance decomposition analysis for included macroeconomic variables but also lead to a complete analysis of the Turkish economy during its inflation-targeting experience in both nominal and real terms through monetary and credit points of view.7 The results show that, although credit card usage has been affected only around 3–4% by the monetary policy shocks (through short-term interest rates), it has been relatively affected more during the implicit inflation-targeting regime (which corresponds to a high level of inflation) compared to the explicit inflation-targeting regime (which corresponds to a low level of inflation), suggesting that low-inflationary regimes are more suitable for a healthier (less volatile) credit card market. Due to relatively higher volatilities in exchange rates during the explicit inflation-targeting period, credit card usage has been affected by exchange rate movements by almost double (around 8%) during that period compared to the implicit inflation-targeting period (around 4%). When it comes to real shocks, credit card usage has been affected by 14%, on the average, during the sample period; when we consider implicit and explicit inflation-targeting periods separately, the former corresponds to an output effect on credit card usage of 18% while the latter corresponds to an effect of 17%. Finally, as expected, the movements in inflation have been more influential on credit card usage during the implicit inflation-targeting period, which corresponds to a higher level of inflation. More important policy implications are on the real side of the economy: During the implicit (respectively, explicit) inflation-targeting period, the movements in output are determined at around 11% (respectively, 5%) through credit card usage, around 8% (respectively, 3%) by the short-term interest rates, another 8% (respectively, 0.5%) by inflation, and 13% (respectively, 1%) by exchange rates. This suggests that output has become self-explanatory during the explicit inflation-targeting regime, which is an indicator of a more stable real economy. This result also has an important policy implication: both the credit view (through credit cards) and the monetary view (through short-term interest rates) seem to be more important during high inflationary episodes for the real side of the economy. When one compares the credit view with the money view, the former seems to be more effective on the real economy during both implicit and explicit inflation-targeting periods (i.e., this last result is independent of the level of inflation). From a policy maker's point of view, this result justifies the motivation of the paper, as depicted above, that there is an increasing necessity for considering the credit channel of monetary policy transmission through credit cards.

This paper has investigated the credit channel of the monetary transmission mechanism through credit card usage in a small open economy, Turkey, which has adopted an implicit inflation-targeting regime in 2002 and shifted to explicit an inflation-targeting regime in 2006. It has been shown that credit card usage has an increasing effect on inflation rates through time; this suggests that there will be a necessity of a monetary policy that will focus more on credit card market in the future. The comparison of implicit and explicit inflation-targeting regimes shows that the output has become more stable, while inflation rate has become less stable and started being affected more by domestic and foreign shocks during the latter. Since the explicit inflation-targeting regime also includes the recent global financial crisis that has started in 2007, these results suggests that the Turkish economy has been mostly exposed to nominal, rather than real, changes during the crisis period. Although this is an indicator of a successfully conducted policy, it raises precautionary concerns for the future as such nominal changes in the current economy may start leading to negative real effects after the crisis period will be over and the global economy will recover. The results of this paper also have an important policy implication: both the credit view (through credit cards) and the monetary view (through short-term interest rates) seem to be more important during high inflationary episodes for the real side of the economy. When one compares the credit view with the money view, the former seems to be more effective on the real economy during both implicit and explicit inflation-targeting periods (i.e., this last result is independent of the level of inflation). From a policy maker's point of view, this suggests that there is an increasing necessity for considering the credit channel of monetary policy transmission through credit cards. Nevertheless, the results of this paper are not without caveats: (i) this paper focuses on the credit channel through credit cards of which movements may not reflect the developments in the overall credit market; (ii) the empirical evidence is from a small open economy; hence, the results may be reflecting some country specifics; (iii) the sample period is restricted to (either implicit or explicit) inflation-targeting period due to the lack of data; comparing the inflation-targeting period with the pre-inflation-targeting period might provide more interesting results for the success of the inflation-targeting regime and its relation with the credit channel; (iv) although the sample period covers most of the recent financial crisis that has started through the end of 2007, the post-crisis period (and, thus, future international shocks through the global economy) is not certain yet; more data are necessary from the future. Focusing on these details is out of the scope of this paper; nevertheless, such extensions, especially analyzing other countries if high-quality monthly data as in this paper will be available, are possible topics of future research.