In the May 21st edition of London’s Evening Standard its financial editor Anthony Hilton laid into the Basel regulatory endorsement of “mark-to-model” risk assessment, citing a speech delivered to the Atlanta Fed by the Bank of England’s Andrew Haldane last month. Hilton reported that:
“In the course of recent research, Haldane has come to some devastating conclusions. First, allowing banks to use their own models to assess the riskiness of their lending — a central plan of the Basel accords — is like letting pupils mark their own exam papers. The results are hopelessly biased towards optimism, so much so that they show things getting less risky for banks when they are in fact getting much more dangerous.
Almost no reliance can be put on these risk-based models, and therefore the entire bank regulatory structure, for another even more fundamental reason. Tests by the Bank of England, the Bank for International Settlements and others have shown that different banks will come to quite astonishingly different conclusions about the riskiness of a given standard portfolio. We are not talking about a vague difference of opinion here but a fundamentally different assessment.”
The thrust of the speech, and of Hilton’s article, was that mark-to-model permitted banks to under-provide against the risk of overvaluation or default on their assets and therefore leave depositors and their government guarantors exposed to unacceptable risk of bank failure. However, there is an additional aspect to this, which leads to reduced stability even when conditions seem favourable for reduced risk and increased optimism. The practice of mark-to-model, and its progenitor mark-to-market, can itself set up the conditions for financial instability.
Banks who carry securities for trading purposes record them on their balance sheets at current market value (mark-to-market) or, if there isn’t an active market, at the value they believe they would be priced at if there had been a market and if that market had set prices in the way that each bank separately believes that it would have done (mark-to model). This means that if markets are rising, or interest rates are falling (which would increase model prices) the value of banks’ trading assets rise and this counts as earnings, even though no sale takes place and no money changes hands. What happens is that the reported value of assets has risen, but there has been no equivalent rise in the value of liabilities, so the difference is added into net worth, as if it were earnings, to make the two sides balance again.
Mark-to-market or mark-to-model creates the basis for a self-perpetuating bubble in financial securities. As market optimism builds up and prices start to rise, banks can book to earnings the rising market value of their assets and thereby accumulate capital, which can be used to support new loans to the finance sector to speculate in the rising markets, driving prices even higher and further enhancing banks’ capital ratios. As company share prices rise but corporate dividends don’t (because these are unaffected by the speculations of the financial markets) long-term investors seeking income will switch into gilts and fixed-interest bonds. The prices of these will start to rise, which means yields (the fixed-sum interest payments relative to the price of the bond) will fall. As yields fall, interest rates start to fall, which increases the value of banks’ mark-to-model securities, which further enhances their capital position, permitting even more lending to finance speculation and, of course, inflating the fraction of capital that can be cashed in to pay bonuses.
Anyone who has wondered how failed banks on life support can deliver consistently rising profits and grant humungous bonus payments in the teeth of an enduring recession now have their answer. Quantitative easing pumped money into the accounts of the investing institutions (pension funds and insurance companies) driving up share prices and bond prices, reducing yields and interest rates and delivering the banks a truly massive windfall on their trading assets, which they promptly cashed in to pay over to their directors and senior staff.
How would this differ under a Positive Money system?
Well we have to work backwards. Firstly (or lastly) quantitative easing was considered necessary because the shut-down in bank lending was starving the economy of money as people and businesses continued (indeed rushed) to pay off their bank loans, and QE was seen as the way to replenish the collapsing stock of money. This wouldn’t have happened under the Positive Money system. The banks, because they would not have had to bear the full risk of default, which would have been shared with investment account holders, would be less reluctant to lend and, even if conditions made people and businesses less eager to borrow, the loan repayments would simply have been recycled to the holders of investment accounts for them to spend instead.
Secondly, the whole regulatory structure which provides such a playing field for banks and their gamester accounting technicians has been erected solely to protect the governments’ guarantees for the savings of people and businesses (the guarantees apply to some non-bank investments as well as deposits). This wouldn’t be necessary under the PM system since the regulatory emphasis would be on educating and informing investors and savers about the true risks which they would themselves have to bear as part of their contracts with the banks and other investment providers.
And thirdly, government guarantees are provided only because of the necessity to secure the nation’s payments systems against banking collapse and therefore are strictly applicable only to the current accounts which feed into those systems. But current accounts are operationally indistinguishable from the time deposits which provide savings vehicles for banks’ customers, so guarantees have to be extended to these also, and these savings accounts are considered to be equivalent to investments held elsewhere in the financial system, so some of these also are included under equivalent guarantees. The whole thing mushrooms. None of this would be necessary under a PM system because transactions accounts, the backbone of the payments system, would not be susceptible to a banking collapse and so would not need guaranteeing.
The upshot is, that under a Positive Money system, however banks decided or were permitted to value their assets on their balance sheets, they would not be able to pay dividends or bonuses or pay-rises, or fit out lavish new premises or display all the signs of obscene wealth, unless they made an actual surplus through buying and selling goods, services (including interest-bearing services) and securities for hard cash.