The procedures and principles governing the repatriation of export proceeds are carried out by the Central Bank of the Republic of Türkiye according to the Exports Circular[1] dated January 16, 2020. An amendment made to the Circular upon the instruction of the Ministry of Treasury and Finance dated December 31, 2021 stipulates that 25% of export proceeds should be sold to banks and transferred to the CBRT's account at the relevant bank. Having been revised to 40% effective from April 18, 2022, this ratio has been implemented as 30% since June 10, 2024. This blog post explores whether firms that sold export proceeds to the CBRT between January 3, 2022 and June 7, 2024 repurchased FX in the spot FX market after the sale, and how the regulations affected their behavior.
Accordingly, we examined approximately 80,000 firms that sold export proceeds at least once in the relevant period and created a weekly data set by combining their export proceeds sales, FX purchases and FX sales. We calculated the net FX repurchase ratio (RR) in the week of the sale of export proceeds using the following formula:
In the equation above, i denotes the firm and t denotes the week. The Market Sale variable represents the amount of FX sales other than export proceeds. Firms are classified as follows based on the value of the repurchase ratio in a given week:
- A negative or zero value (RR≤0) denotes the firms that did not make any repurchases on a net basis,
- A value between zero and one (0<RR<1) denotes the firms that made partial repurchases,
- A value equal to one or above (RR≥1) denotes the firms that made repurchases that equaled or exceeded the amount of export proceeds sales.[2]
Findings from the calculations based on this classification indicate that 79% of the firms that sold export proceeds in the given time interval did not make any repurchases in the week of the sale of export proceeds, 12% made repurchases of less than the amount of the sale of export proceeds, and 9% repurchased the entire amount of export proceeds sales, or more (Chart 1). In terms of the amount of export proceeds sales, these ratios correspond to 67%, 21% and 12%, respectively. In other words, 67% of the total amount of export proceeds sales accounts for the firms that did not make any repurchases during the week of export proceeds sales. Moreover, an assessment of all export proceeds sales of the relevant period together with spot market transactions reveals that 74% of the firms that repurchased all or more of their export proceeds that they sold during the given period were net importers. Therefore, findings suggest that firms selling export proceeds were mostly not eager to repurchase FX, and the majority of the firms that repurchased FX were net importers.
On the other hand, the additional exchange rate difference burden that firms incur after selling to the CBRT and repurchasing in the spot market on the same day in the same currency, at any amount, and at an exchange rate higher than that of export proceeds sales is estimated to be approximately 8 per thousand on average. When the wide exchange rate buying-selling spread in the interbank market during the first half of 2023 is excluded, the exchange rate differences become more reasonable.
In addition, our study also includes a review of the regulation,[3] which took effect on January 26, 2023 and introduced a “2% exchange rate difference payment in return for a pledge not to buy FX equal to the amount of FX sold.” To this end, we analyzed the impact of this regulation on the probability of firms being net sellers of FX by using linear probability models[4] with firm-level microdata. The analysis controls for firm heterogeneity and time (weekly) fixed effects, taking into account that different firms may have different FX needs.
According to the results presented in Table-1:
- On average, the probability of being net seller of FX is 51% higher for firms that sell export proceeds in the relevant week compared to other firms in the sample.[5]
- Raising the minimum ratio for sale of export proceeds from 25% to 40% increases this probability by 5 percentage points. In other words, the regulation demanding that firms sell 40% of export proceeds to the CBRT increased the probability that such firms would be net sellers of FX in the week of the sale.
- The regulation of 2% exchange rate difference payment in return for a pledge not to purchase FX equal to the amount of FX sold after the sale of 40% of the FX obtained from abroad to the CBRT also slightly increases the probability of firms selling export proceeds to be net sellers of FX.
- FX sales to the market by firms selling export proceeds other than the amount they have to sell to the CBRT are significant in terms of liquidity. Firms selling export proceeds are also 11% more likely to make sales to the spot market than other firms included in the sample. However, this ratio drops by 6 percentage points after the enforcement of the 2% exchange rate difference payment regulation. This implies that some firms prefer to sell to the CBRT some of the FX they have already consented to sell to the market due to the 2% exchange rate difference payments.
In sum, this blog post closely examines the effects of the implementation regarding export proceeds sales to the CBRT, which entered into force on January 3, 2022, and of the subsequent revisions to that policy on firms' FX repurchase behavior and the FX market. The findings suggest that the vast majority of firms selling export proceeds do not repurchase FX on a net basis. Analyses based on firm-level microdata indicate that firms selling export proceeds are more likely to be net sellers of FX in the week of export proceeds sales when compared to other firms, and that the related regulations have increased this probability.
[1] For the most up-to-date version of the Exports Circular dated January 16, 2020, which regulates the procedures and principles regarding the repatriation of export proceeds, see https://tcmb.tl/wj6bac6 (in Turkish only)
[2] Classifications are based on net FX purchases.
[3] For the press release on the 2% exchange rate difference payments, see https://tcmb.tl/wi61ea5
[4] The regression models are run by the equation 
The variable Yi,t is a binary (1/0) representation of each firm's status of being a net seller of FX in the relevant week in model (1), and of being a seller of FX exceeding the minimum amount of sale of export proceeds in the relevant week in model (2). The variable Xi,t denotes each firm's status of selling export proceeds in the relevant week in binary form. The variable POST1 takes the value 1 after April 18, 2022 and 0 before April 18, 2022, while the variable POST2 takes the value 1 after January 26, 2023 and 0 before January 26, 2023. di controls for firm-fixed effects, and dt controls for time (week)-fixed effects. The variable ϵi,t denotes the error term.
[5] Developments in exchange rate expectations and domestic interest rates can be expected to influence the probability of exporting firms becoming FX sellers. To control for this effect, the analyses employ time-fixed effects. In addition, to test the robustness of the findings, separate regressions are run for two different periods when expectations are relatively negative (January-April 2022) and positive (January-June 2024). The results reveal that firms selling export proceeds are more likely to be net sellers of FX in both periods compared to other firms in the sample. Meanwhile, the estimated effect is even higher in the period when expectations are relatively positive compared to the period when expectations are relatively negative.