A brief review of the economics of financial market intervention

The storm over end 2009 bonuses

What are the social costs of the measures put in place ?

State guarantees for bank debt

Guarantees to banks will only cost money if the bank cannot subsequently honor its debts. If the bank pulls through, the state will pay nothing. State guarantees have generally not been provided free. Banks have effectively won access to liquidity via money markets in exchange for giving the government capital. The social cost of this intervention is the difference between the amounts paid out to honor guarantees discounted to the time of sale of the capital and the value recovered by the sale of the shares. As the share have generally been preferred shares, we can assume it will be the bank itself (or its merger partner) who will pay for the preferred shares, probably issuing ordinary shares at the same time. Generally, governments will hold a collateral value for the amount of the guarantee as tap stock for the period of their intervention. This means treasuries will be bought (or retained from a new issue) by the central bank for the account of the treasury. This is an addition to government debt issued. The final cost is not known until the capital participation in bank capital is unwound. If the bank earns profits and pays dividends in the mean time, the state will receive a dividend as a shareholder.

Direct capital participation

The government may simply buy a chunk of a company, boosting its cash position to purchase ordinary or preferred shares. First there is the question of the context in which such a transaction may occur. Such an intervention will usually take place while the value of shares in general and the company concerned in particular are crashing so fast that the capital-value of the company can no longer support its liabilities. The aim is therefore to inject liquidity and to prevent the capital value falling further. To this end, the government would most effectively act as a “buyer of last resort” buying all shares on offer at a certain level in the market, where this level would be at or above the level reached. The social cost of this intervention is the discounted value of the difference between the price paid and that received, when the shares are resold. In the case of final insolvency, the amount would be lost. The other comments made above for the guarantee are applicable here.

Depositor guarantees

This is an amount paid directly to bank’s retail clients if the bank cannot honor its obligations to them. It is usually capped at a certain amount and only banks which partake in deposit insurance schemes are covered, but as these schemes are mandatory in EEA-countries, all banks in these countries are covered by at least the mandatory scheme. Guarantee payouts to retail bank depositors are a straight cost in terms of government expenditure. These require either a reallocation of spending resources or higher taxes or more debt (or a mixture of these). The funds paid out may be used to finance consumption, finance other investment or be re-invested in treasuries. In terms of the social impact, there is certainly a potential reallocation from one group to another ; the depositor receives the cash and not the defense department or the state kinder-garden system. If we assume the deposit insurance amount paid out is reinvested by recipients in safe assets (treasury bonds), then these will finance their own insurance payout. [This need not be direct, they could buy a safe money-market/bond fund and this goes and buys treasuries]. If such intervention is financed by a new treasury emission, or the sale of tap stock, the economic impact in this case will be delayed until the treasury expires. The government now pays the coupon instead of the bank paying interest. Also, the additional debt issued may crowd out other assets and raise the cost of long term borrowing. However we may ignore this factor, as the assets being replaced are those whose non-performance presumably resulted in the demise of the bank in the first place. Alternatively, the amount could be reinvested in other riskier assets or consumed. The first alternative seems intuitively unlikely during a financial crisis, but it is not impossible. This additional investment demand for capital assets would tend to support asset prices. Thus a potentially stabilizing impact on financial markets could be expected. Additional spending financed by a government payout would have a directly reflationary impact. Presumably also a welcome impact in a financial and economic crisis.

Direct government lending to banks

An unsecured loan to a bank will result in a straight loss for the government, if the bank subsequently goes bankrupt and nothing can be recovered. In fact, banks must provide collateral assets to the government of sufficient value and quality to cover the loan. The loan would be executed by the central bank in the same way that it executes other open market operations. In other words, we would normally expect the central bank to execute a reverse repo with the bank. In the case of a bank default, the conditions of the loan should ensure that the central bank may recover and liquidate sufficient assets to cover the loan. There is a market risk involved in this, but is the loss to the government should be close to zero. The social cost of such a support measure should be no greater than any other central bank intervention.

Low central bank interest rates

The central banks in many countries are holding their intervention and lending rates low to help instill confidence in financial markets and to provide banks with a margin, supporting profits and helping to rebuild reserves. This policy was practiced for many years by the Bank of Japan in the 1990’s after the Japanese bubble economy burst. Looking back at this period we can say that the Japanese banking system managed to consolidate, without crashing the economy. During this period unemployment remained relatively low and despite the crash in asset prices, the recession was largely limited to a growth recession – albeit, a long, drawn out one. There may be some parallels for the present crisis : By reducing the pain for the banking system (internationally), the restructuring of corporate entities is likely to progress more slowly ; the impact on the economy of bank restructuring will be softer. Until we know what the impact on the economy of bank restricting will be, it is impossible to say whether going slowly is good or bad. One would hope (naively ?) that bank restructuring and changes to business models would be for the better, in which case delaying these changes would have to be viewed as an economic loss. Certainly, there is a point when the availability of a resource (central bank liquidity) at prices close to zero disengages the mechanism of price as a factor or resource allocation.

What are the lags between rebuilding bank capital and increased bank lending ?

This is a point reminiscent of the discussion of intervention in the savings and loans crisis at the beginning of the 1990s. The crisis had sapped confidence in borrower quality and banks were unwilling to lend to clients (neither corporate nor household). At the time, central bankers complained that promoting bank lending by keeping central bank intervention rates low was like “pushing a piece of string”. In the 1990’s Japanese banking crisis, low central bank lending rates did not make it easier for companies or individuals to get credit. This complaint is reflected in the current crisis, where both central bank and treasury officials are complaining that the low interest rates targeted by central banks are not being passed onto the economy as lower interest rates. (See the article for further discussion of the constraints of macro prudential instruments – Central bank policy and asset inflation- comments on the IMF WEO October 2009 “Lessons for Monetary Policy from Asset Price Fluctuations”). On the face of it, tighter definitions of bank capital and stricter application of capital adequacy ratios will reduce bank lending. In theory we can identify a price effect on lending (low price = higher demand), a multiplier effect (capital and reserve ratios determine the volume lent) and leads and lags (behavior influenced in advance by expectations, delayed decisions due to uncertainty). There are a series of lags involved in the realization of a more expansive business model in banking. To some extent, these are imbedded in the institutional structure of banks. To some extent these lags are caused externally, as banks wait to see how regulatory changes pan out before committing to new business models or simply to new business.

Institutional lags

Credit policy will be orientated to only accepting good risks on to the books when times are uncertain. Access to cheap funding and the promise of big profit margins will quickly lead traders to exploit arbitrage possibilities. Profits at the end of 2009 showed that the investment banks smart enough to survive 2009 could certainly earn money in such conditions. However, these same conditions will not inspire credit departments to go and lend money to people they think are not very likely to pay it back. This is especially so in an economic crisis where corporate clients face shrinking order books and retail clients face increased risk of unemployment. The point of departure for the credit department in a crisis is that, even if the potential to earn 1% more margin exists, this is not much incentive, if 20% of the loan portfolio is classed as doubtful or non-performing. Only when the economic outlook of borrowers improves can we expect banks to start lending more freely. This is a vicious (or virtuous) circle is only marginally affected by central bank lending rates. One of the trends of the last 15 years or so has been towards securitization of debt. This permits the lending bank to either resell the loans to reduce the impact on the balance sheet of this business. Alternatively, the credit risk associated with such loans may by reduced synthetically by using a similar structure, but buying CDS instead of selling the loan portfolio. Various changes are being discussed in the regulatory environment, both for the definition of capital and for capital ratios. These are key factors a business model of originating, selling and holding certain components of securitized debt to finance lending. We may therefore expect further institutional lags to generating more lending, while the regulatory environment is reassessed.

Assessing structural market weaknesses

Quite apart from the impact proposed regulatory changes will have on existing business models, the existing models have demonstrated some notable weaknesses during the crisis. We may expect a certain reluctance to greatly expand lending covered by securitization in sectors where sizable losses have been made. This is likely to be the case where underlying asset markets are illiquid, or where credit risks can no longer be adequately covered by swaps. The derivatives markets proved a key risk factor during the crisis and it is consequently receiving attention both from regulators and from banks. Regulators are likely to propose measures assuring a greater correspondence between short derivative positions and corresponding hedge positions in the underlying cash markets. , lack of correspondence between derivatives and underlying. This may well have a negative impact on the liquidity and on spreads in some markets, possibly making business models based on them potentially less competitive in terms of pricing. Steps to redefine capital and review capital and liquidity ratios discussed above may have an additional slowing impact on lending based on such models. Low liquidity for underlying cash markets will remain variable, but one might expect lending to recover faster in sectors with either more liquid underlying markets or with better established legal rules pertaining to default and liquidation. OTC derivatives proved an effective method of contagion during the crisis, as Lehman Brothers was a major player in these markets. The failure of Lehman was a key element in spreading the crisis, as AIG was left exposed with many positions in CDS for which it had no hedge. Banks, clearing houses and banking associations have already started to use central counterparty clearing for some OTC derivative settlements. Provided the CCP is sufficiently collateralized by the participating parties, the wider application of this existing solution offers to give better protection from contagion to banks. As improved settlement solutions for derivatives are realized by the market, we can expect a positive impact on volumes and possibly on pricing. This may expand the possibility to use synthetic structures in loan securitization, with a positive impact on volumes. Indirectly this may help a recovery in lending volumes.

Alexander BLINKHORN