Liquidity and Market Failure: lessons from 1994

by George Hatjoullis

In the summer of 2013 there was widespread talk of a bond market crash and I wrote a brief blog highlighting the lessons of the bond market crash of 1994 (http://bit.ly/1GTKjza). My main concern was to show that rising yields did not necessarily imply a poor performance from equities. Over the years 1994-1995, the US 2 year Treasury note yield rose from 4% to 7%. Over the same period the S&P 500 rose 30%+. I conclude:

Indeed, if one studies the… charts above one might conclude that, apart from very recently, rising yields tend to coincide with increases in the S&P 500.

My purpose was to caution extrapolation from bond markets to equity markets. In the event both rallied. My career was based in interest rate products and currency markets, yet in retirement I have focused exclusively on equities. In part it is a desire to learn something new but mainly because only equities offer real potential for portfolio growth. Like many retirees in defined contribution schemes this is key to a comfortable retirement. In this blog I return to my roots and look at the current state of the bond markets.

I was motivated to do this by another insightful article from Gillian Tett (my favourite FT journalist). In this article (http://on.ft.com/1dHiKiQ) GT highlights the poor state of liquidity in bond markets at present and how it is a direct consequence of prudential regulatory policy. She rightly expresses concern that, in a rising interest rate environment, an attempt to sell bonds to mitigate interest rate risk might overwhelm a much diminished market making and online broking mechanism and cause a market failure. As I highlighted in my letter to the FT in response to the GT article (http://on.ft.com/1E8Kp1M), this is precisely what happened in 1994. On February 4, 1994, the Federal Reserve raised interest rates by 25 bp. The resulting avalanche of selling overwhelmed the then very liquid bond market making structure and caused a market failure.

The situation in 1994 was however quite different to the present and the purpose of my letter was to highlight this difference. In 1994 the distinction between market making and proprietary trading was a little blurred (I worked in this blurred area so this is an eyewitness account). When the client selling began, the market making institutions were also overly long bonds and interest rate products and did not want to receive. In the time-honoured tradition, they bid-to-miss (so to speak), which widened bid/offer spreads and pushed prices lower. The markets were one-way. There was no bid really. This was bad enough but it is  at time like this that all the sins of the past get flushed out (as happened in 2008).

In the period between 1986 and 1994 a whole new panoply of structured products had emerged and been pushed by avaricious salesmen on unsuspecting clients. These products often had what the US mortgage market (yes this monster again) would call negative convexity. One product that had become ubiquitous was the yield curve note. This bond paid a coupon that varied inversely with the short-term interest rate. The result was that its price volatility with respect to short-term interest rates was about twice that of a fixed coupon bond of the same maturity ( It could also be structured to be even more volatile and it was possible to leverage the underlying note). Moreover, the price goes up as short-term interest rates rise unlike the price of a fixed rate bond. It was designed as a hedging instrument for those having natural fixed coupon portfolios. However, being a structured product it was very profitable to issue and so was also sold as a speculative instrument.

The institutions that issued these notes had the opposite position to the client base. They were collectively short something that increased their long interest rate exposure twice as fast (or faster) as interest rates rose and so found themselves selling short dated bonds and derivative equivalents into a falling market. This created a self-defeating cascade. Moreover, the presence of such structured products effectively leveraged the market and in a non transparent way. In addition, there was huge leverage of a more obvious kind from hedge funds and banks alike. Yields had fallen continuously since 1990,in part because of the savings and loans crisis in the US (sound familiar?), and this had encouraged daft speculative leverage. In 1993 it was particularly noticeable. This leverage tried to unwind as one sale, after February 1994. The result was a market failure that left an indelible mark on the psyche of anyone that experienced it.

The reason for outlining this period is to emphasise that important differences may influence the course of events as the Federal Reserve raises interest rates. First, whilst the regulatory framework has reduced bond market liquidity, it has also reduced leverage and the proliferation of inappropriate structured products. Secondly, the motive for bond longs today is quite different to the motive in 1994. In the latter case it was entirely speculative. Today it is largely defensive. As previous blogs have annoyingly repeated, the defining characteristic of 2008 was the demise of the too-big-to-fail view of banks rendering uninsured deposits risky. The holding of government bonds today is partly to avoid the credit risk now associated with uninsured bank deposits. In 1994 it was to benefit from interest rate risk. A portfolio of high-grade corporate bonds offers a better yield and arguably is no more risky than an uninsured deposit, albeit less liquid (thank you Andres Drobny for reminding me to clarify). A rising interest rate environment does increase the interest rate risk from holding fixed coupon bonds. However, credit risk is still a major consideration and interest rate risk may be deemed to be preferable to credit risk. Moreover the latter may also rise as interest rates rise. There will be bond selling in response to increases in short-term interest rates but, despite the poor liquidity, a repeat of 1994 is not inevitable. Finally, the credit risk concern may encourage holders of high-grade bonds not to sell and to use the derivative markets in order to hedge out interest rate risk. The poor liquidity in bond markets may thus not be an immediate issue. This could, of course, lead to a dislocation between cash and derivative markets but the latter remain very liquid so there is no automatic reason to expect market failure. Still, it is worth keeping a wary eye and one hopes the regulators are doing just that. I should like to add that if I am aware surely they are but regrettably this was not true in 1994 or 2008.

Postscript 7/05/2015

The long-awaited bond market correction has begun. The timing may be seasonal but I have been made aware of one aspect that was not evident in 1994; the systematic fund universe. These funds are automatic trend following systems. This would explain why yields managed to get so low. It also raises some more concern about liquidity. These funds sell on signals and do not have emotions or thoughts. They also all tend to do so at the same time. Moreover, they sell the product that they have bought. It is possible that such selling could overwhelm the market in the short-term and cause some serious indigestion. It is unwise to enter the market on the long side until these funds have stabilised their activity.

Advertisements