Turkey: To The IMF Again!

Publication date
Monday, 12.11.2001

Augusto Lopez-Claros Tolga Ediz



Turkey faces a difficult year ahead, having failed in three previous attempts at macroeconomic stabilisation in the last two years. The fourth attempt will be under the shadow of ugly macroeconomic fundamentals:

  • A worsening growth environment poses a major risk to financial stability.
  • Debt rollovers remain onerous next year, with an alarming rise in the share of FX-denominated debt.
  • Structural reform implementation will take place under weakened societal consensus.

Given these challenges, it is not surprising that Turkey has once again knocked on the IMF’s door. In our view, if “strategic considerations” continue to guide aid to Turkey it would be more appropriate for the G7 to step in with bilateral (ie, reschedulable) funds. 

Worsening Fundamentals

IMF credits are given in support of economic reform efforts. Quick-disbursing loans are supposed to give the authorities room for manoeuvre as they take the necessary measures to put the economy back on a path of non-inflationary growth that enhances the country’s debt-carrying capacity. The logic of Fund conditionality dictates that there be a large and positive correlation between the amount of foreign financing being provided to the country and the scope of the underlying reforms, to generate the resources that will maintain the revolving nature of IMF resources. Against the targets of the three-year programme approved in December 1999, performance in Turkey has been abysmal. Indeed, the macroeconomic fundamentals have deteriorated in a major way. Inflation (now running at 80%) is actually higher than at the programme’s outset, while the projected debt stock for end-2001, at 90% of GNP, is more than 33 percentage points higher than originally envisaged. The lira which, by now, was supposed to have been enjoying the fruits of a prolonged period of stability, has collapsed and is likely to remain under considerable downward pressure.

Notwithstanding substantial progress on the structural front, the authorities’ programme is threatened by a rapidly worsening economy, which is in a much deeper recession than most analysts and the authorities had expected. Even our own March forecast of a 7.5% real GNP contraction in 2001 (far more pessimistic then than the 3% drop contained in the May IMF programme) may turn out to have been on the rosy side. Recent data suggest that a 9% GNP drop this year is more likely, and the outlook for next year is bleak. Against the background of a weakening global economy, Turkey will find it tough to attract much-needed foreign capital to restore growth, including raising funds in the eurobond market. The authorities’ hopes that the programme would precipitate an export-led recovery have been dashed in a major way. Such a sharp contraction poses a number of challenges.

The IMF programme: targets and projected outturn

2001 1 2001 2
3-year standby Outturn
Real GNP growth 5.2 -9.0
Inflation rate, eop 12.0 73.0

Public sector debt, % of GNP

56.6 90.0
Primary surplus, % of GNP 3.7 5.0
Lira/$ rate, eop, 000s 775 1750

1.    Targets contained in 3-year standby (2000-02) approved on 12/1999.

2.    LBGE forecasts.

First, the government will find it increasingly difficult to continue to pursue credible fiscal adjustment. The authorities are aiming to achieve a primary surplus of 6.5% of GNP in 2002, up fr om 5% this year. But the clampdown on non-interest spending implied by these targets - difficult under any circumstances - will be especially difficult in the context of a contracting economy engaged in a third year of belt-tightening.  A weak economy will inevitably add to the costs of banking sector restructuring, as it did in a huge way in Indonesia.

Second, investors’ concerns over the health of the banking system and the scale of the emerging private-sector debt overhang are intensifying. Many corporations are in deep financial trouble, having suffered major FX losses after the collapse of the exchange rate peg last February. The ensuing sharp rise in interest rates and the drop in profits have led to further debt-servicing problems. Banks’ non-performing loans (NPLs) have risen dramatically, perhaps to as much as 25% of total loans outstanding. Moreover, as a result of banks’ more cautious lending, a “credit crunch” has prolonged the economic downturn.

Third, lower growth could undermine what little societal and political consensus remains on economic reform. As we have argued, some consensus among the social partners about the objectives and instruments of the programme, as well as the distribution of the costs, is crucial for the success of any stabilisation effort. There was little evidence of this in Turkey from the outset. Indeed, it was never clear how the present government - with its low levels of public support - could make a credible case for painful economic reforms.[1] Worsening growth prospects have since made this case even more difficult.

Debt Dynamics In 2002

The rollover of debt obligations remains the overriding short-term macroeconomic problem. We have upd ated our estimates of financing requirements for next year (table), assuming that, in the new environment, only $2bn will be available from eurobond placements.[2] No allowance is made for additional IMF funds beyond the levels envisaged in the present programme, nor do we prejudge the outcome of any “restructuring” requests from the Turkish government on about $6.5bn of debt to the IMF falling due in 2002.

Public sector financing needs in 2002
(TL quadrillion)

Domestic debt service 134.8
   Market debt 74.8
   Non-market debt 60.0
External debt service 24.3
Total debt service 159.2
Sources of finance 159.2
   Primary surplus 18.2
   State bank debt rollover 17.1
   Deferred payments to central bank 11.1
   Eurobond issuance 4.0
   IMF 2.3
   World Bank 5.0
   Other borrowing 9.0
   Auctions 92.5
Memo items:  
Nominal GDP 280
GDP in $bn 140
FX denominated/linked domestic debt 22.0
Auctions as % of GNP 33
                     % of deposit base 60
                     % of TL deposit base 149

Source: Turkish Treasury, IIF, and LBGE estimates.

Our analysis suggests that debt rollovers next year will remain extremely difficult. First, in the absence of additional external support, the treasury would have to issue around TL 92 quadrillion of debt to the private sector (33% of GNP), a figure that is large in relation to the contribution made by the private sector to total GNP. Put another way, debt issues to the private sector could take up to 60% of the total deposit base, confirming our long-held premise that the financial system in Turkey is not deep enough to finance government borrowing. A nearly broken banking system would be stretched to the lim it to roll over domestic debt, resulting in continued contraction of private-sector credit extension. Such massive crowding out of the private sector would drastically reduce hopes of a resumption of growth.

Furthermore, the amount of TL debt to be issued to the private sector is more than three times the size of the Turkish lira deposit base of the private-sector banks. Debt rollovers next year are virtually impossible unless a solution is found to this “mismatch” problem, although the transfer of deposits from SDIF banks to private banks would help a bit. There are at least three ways to address this: (i) banks could increase FX risk and use their FX deposits to rollover debt; (ii) the treasury could issue FX debt to the private sector, reducing its reliance on TL denominated resources; (iii) dollarisation could fall as depositors and corporations switch back into TL assets.

The first solution would appear to require a return to previous “hot money” policies and would involve a major increase in banking sector risk; this is unlikely to happen in the current environment. The second option, involving the issuance of FX debt, has been used extensively this year, resulting in an alarming increase in FX-denominated debt and raising questions in the minds of investors about a rapidly emerging FX debt overhang, dangerous given the continued pressures on the exchange rate. By ruling out monetisation as an option to reduce the domestic debt stock, the authorities have opted for a rapid change in the currency composition of the public debt, a high-risk strategy with long-term implications for the budget and the ability of future governments to service FX obligations on time.

Reforms In 2002: All About Implementation

A return of confidence, leading to a reversal of currency substitution in favor of the lira, is essential. At a minimum, this would require credible implementation of a considerably enhanced programme involving major headway in the dismantling of the rent-seeking mechanisms presently at the centre of public-sector management. This approach would involve several features. First, enactment and immediate implementation of the state tender law, including the appointment of an independent arbitrator for awarding government contracts.  In this respect, recent hints that the government might be tempted to postpone the enforcement of the law until 2004 - that is, until after the elections - would be a major, possibly fatal, blow to the programme, particularly as regards the regaining of confidence on the part of foreign investors.

Second, now that the state banks are se t to resume lending to the corporate sector, it will be important that this be done free of the interference and insider dealing that has come to be the hallmark of these institutions. The independence of their management has to be nurtured and credible evidence must emerge that they are no longer the fiefdoms of corrupt politicians. Third, with the continuing cleanup campaign in the state sector, the government should not hesitate to cancel expensive energy contracts negotiated on criteria other than efficiency. The social costs associated with all these measures are negligible, but the potential payoffs in terms of the signals that they would send about improved future fiscal management are enormous.

Fourth, the authorities have much further to go in bringing about a permanent improvement in public finances through major restructuring of state enterprises. It may be “politically unrealistic” to suggest that a government held in low regard by the public will have the political will to reduce the size of the public sector through increased use of early retirement schemes, the closing down of non-viable operations, and a shift of public spending away from pet projects - including in the military - towards education, health and infrastructure. But, the reality is that there is no alternative. These things will have to be done by this or any other government and political expediency, as a short-term strategy, is likely to be lethal for those politicians practising it.

The IMF Again

Given the above challenges, it is not surprising that the authorities have, for the fourth time in less than two years, turned to the IMF for help. For sure, Turkey’s short-term debt dynamics would be less ugly if additional external funding were to be provided in 2002. For example, debt issuance to the private sector could fall by as much as 10% of GNP if the IMF were to approve a $14bn package next year, consisting of $4bn of refinancing under the Supplementary Reserve Facility and $10bn of fresh money. As long as the funds were used for domestic debt refinancing, IMF money could reduce the rollover requirements next year and ease the crowding-out of the private sector. But rollover requirements would remain large, with more than 20% of GNP still needed to be issued. Hence, IMF funds would not necessarily be sufficient to restore confidence.

In our view the scope for additional IMF funding in 2002 has been exhausted. Confidence has not been restored, debt levels have skyrocketed and the future claims on budgetary resources that these represent raise serious questions about external sustainability. We are also concerned that Turkey’s requests for new funding, a mere five months into the implementation of one of the Fund’s largest programmes ever are, once again, being cast in “strategic” terms, having mainly to do with Turkey’s role in the war against terrorism. Indeed, the only reason why such requests have not been summarily dismissed by donors is that Turkey’s ability to deliver on the security front is linked to the authorities’ ability to forestall the onset of yet another financial crisis, with unpredictable political ramifications.

The problem, of course, is that, beyond a short-term rally, financial markets are unlikely to be impressed by such considerations. Turkey’s future economic recovery depends on the continued implementation of credible structural reforms, not on how accommodating its government is in supporting the goals of the coalition. We are thus of the view that, moving forward, “strategic lending” should be done on a bilateral basis, through the provision of credits, the disbursement of which could be made conditional on satisfactory compliance with respect to strengthened IMF programme conditionality. The advantages of this approach are manifold. First, there is already a framework in place for the restructuring of official debts; in due course G7 creditors can agree to sit down with the Turkish treasury to discuss, as many times as needed, future reschedulings in the context of the Paris Club. This is quite important since we expect the Fund to be engaged in Turkey for many years in “refinancing” operations; putting a cap on debt to the Fund would make future debt management a tad less onerous.

Second, G7 creditors could avoid discrediting a little less the reputation of the IMF for even-handedness of treatment across its 183 members. Were Turkey’s debt to the IMF rise to $30bn by end-2002 (a scenario fully consistent with the market expectations in Istanbul and Ankara), then Turkey would account for about 40% of the total debt of all the Fund’s members – a sobering statistic. The global economy very much needs the IMF, but a minimum of credibility is vital for the organisation to be effective. Finally, with closer G7 involvement in funding and policy formulation, the scope for a possible improvement in confidence might be enhanced. There is little doubt that, say, a $10bn G7 credit would be more bullish for the market than a fourth IMF bailout of the same amount.


Given that “strategic considerations” are likely to continue to guide aid to Turkey in 2002, it would be more appropriate for the G7 to step in and provide bilateral support; $20 billion would help. But, as happened last May, the IMF may, once again, be forced to make a potentially large financial contribution to this effort.

[1] See: “Turkey: Five Key Lessons”, Global Weekly Economic Monitor, 9 March 2001.

[2] We assume that bond yields fall to 70% (from 85% currently) and that the treasury rolls over much of the debt principal held by state banks.  For details on the methodology, see “Turkey: Walking On A Tightrope” by Tolga Ediz and Kaushik Rudra, Sovereign Strategist, 7 September 2001.



Go to other releases