Although not at this week’s Pension PlayPen lunch, Con asked for an account of the conversation; I mentioned that those in the room were pleased that schemes owning large amounts of index-linked gilts, had enjoyed bumper returns last year. It is worth reading to the end of this great shaggy dog of a blog to understand the import of its title
In late 1978, after seemingly interminable discussions with lawyers and accountants, the late Chris Golden and I called on the NY Federal Reserve; the purpose was to request that the Fed facilitate the stripping of US treasury bonds, which was at that point a jury-rigged, Heath Robinson process. They were intrigued that representatives from two competing government securities dealers, and foreigners at that, should find common cause in this. Their questions came thick and fast.
They had known for some time of tax motivated stripping; coupons were subject to both a withholding tax and income taxes, while the principal was simply a matter of capital gains tax; they were not inclined to encourage this. We explained that we wished to resolve a corporate finance issue, and that in doing so, this would create a new source of stable demand for treasury securities.
At that time, many US corporations had numerous small legacy bond issues outstanding, which were quoted at deep discounts to par value. It made sense to retire these, but purchasing them in the market was not a viable course of action. Buying these securities at a discount to book value generated taxable profits for the corporations. However, few investors were comfortable realising losses, and many owners could not even be identified.
The solution lay in a process where the cash flows of these outstanding securities were matched by treasury instruments lodged, in escrow, with the bonds’ trustees or fiscal agents; dedicated portfolios. The process was termed defeasement. Legacy issues would no longer appear on the corporate balance sheet, and the difference between the cost of the dedicated treasury portfolio and the book value of securities retired would be posted as income. There was even talk that credit ratings might be improved.
This arrangement would be a new source of particularly attractive demand for the Treasury as the securities would be held to maturity.
The Fed demurred and did not formally introduce STRIPS (Separate Trading of Interest and Principal) until February 1985, though by then, with withholding taxes removed, many proprietary work-arounds had been created; and numerous corporations had retired legacy issues, having defeased them in the Volcker bond slump of 1981-82.
In a manner prescient of the current market in index-linked gilts, the STRIPS market has from time to time traded anomalously as a result of defeasement activities. US state and local government agencies used them extensively in the early 1990s. The demand made the process of stripping very profitable for government dealers. The Brady bond programme was another source of demand at that time. The return of much of this Brady collateral to the market, after the global capital markets reopened to sovereign borrowers around the millennium, would have proved far more problematic were it not that the Fed was, at that time, repurchasing governments.
The most important question raised in that initial Fed meeting was: why would anyone wish to match assets and liabilities? This remains the critical question for liability driven investment in the context of DB pensions. There is no compelling theoretical support for such an approach for pension scheme management, rather it is well-known that such dedication is among the most expensive forms of management.
One argument advanced is that pensions are “bond-like”. They are, but only in the very strict sense that they are fixed claims, meaning that they do not participate in the upside of equity. However, they are not “bond-like” in the sense that interest rates do not figure in their construction; neither the contributions nor the benefits payable vary with changes of interest rates. There is an implicit contractual accrual rate defined by the contribution and projected benefits amounts, but interest rates are not among the risk factors of pension schemes. Modified duration, the standard risk measure of bond analysis and risk management, where it can be estimated, is spurious and irrelevant.
By contrast, the Pension Protection Fund does have exposure, and sensitivity to interest rates, which is introduced by the manner in which they price and acquire pension liabilities. The mistake here is that the Pension Regulator is applying risk measures which are appropriate for the PPF to open ongoing schemes, where they aren’t. The 2005 scheme funding regulations compound this confusion by specifying forms by which discount rates may be determined.
Market consistent accounting standards, again a bond basis, are subject to the Duhem-Quine critique that claims (or objectives) are unable to be confirmed or falsified on their own, in isolation from surrounding hypotheses; in this case, the hypothesis is the fundamental theorem of asset pricing. This may be stated as:
- A market admits no arbitrage, if and only if, the market has a martingale measure.
- The martingale measure is unique, if and only if, every contingent claim can be hedged.
These are the conditions which unify the theoretical frameworks of regulation, accounting and fund management practice. They are clearly not met in the world in which we live.
The result has been that regulation and accounting have driven management practices which are both irrational and incredibly costly – LDI and de-risking. Investment descends into price driven speculation. The present price levels of index-linked gilts illustrate this perfectly; why should anyone want to save into an instrument which guarantees them lower future purchasing power?
The self-fulfilling nature, the performativity, of model driven activity is well known in practical finance. Prior to its widespread adoption, the prices and hedging ratios of the Black Scholes model of options were far from those observed in markets; in those early days, Fisher Black lost a fortune as a result. By the mid-1980s, with widespread adoption, the market behaved very much as the model implied; until one day of crash in 1987, after which it no longer did. Even pit traders can learn. There are also negative forms of performativity; greater fool theory is one, the pursuit of momentum or trends regardless of fundamental value.
Those congratulating themselves on having the acumen to have owned index-linked gilts last year should remember this. With respect to the fundamentals of pensions, interest rate movements are sunspots, an extrinsic random variable. Irrelevant. However, expecting this to end in cathartic tears is ill-conceived; we have been largely drained by the pain, cost and mental anguish already experienced.