IT MUST BE tough to be fingered as one of the main culprits for the biggest stock market crash in history. Hayne Leland appears to have borne the experience well. 'I [went] from being somewhat of a hero to somewhat of a goat,' the famed finance professor said, smiling wryly as he relaxes in the hotel lounge. Mr Leland invented 'portfolio insurance', blamed by the Brady report into the 1987 stock market crash for causing a self-perpetuating downward spiral of selling on Black Monday, October 19. The Dow Jones industrial average fell 22.62 per cent or 508 points that day, a record one-day decline. It was an ignominious reversal for a brilliantly simple investment strategy that for years had worked flawlessly. Investors have long searched for the holy grail of investing: a way of making money when prices rise without running the risk of losing when they fall. Mr Leland seemed to have discovered the answer, hitting on the idea during a sleepless night in September 1976, according to the book Capital Ideas by Peter Bernstein. Puzzling over how to insure a portfolio against declines, Mr Leland, then a 35-year-old professor at the University of California, Berkeley, realised that a put option worked identically to an insurance policy. For a small premium, it protected the investor against a fall in an asset's value. A put option gives the holder the right to sell an asset at a certain price. Its value rises as the asset's value falls. If an investor buys a put option on his portfolio, therefore, any drop in its value is offset by a corresponding gain in the put option. However, in 1976 there was no organised market for options on stocks, let alone whole portfolios. Drawing on the work of Fischer Black and Myron Scholes, the pioneers of option-pricing theory, Mr Leland worked out you could create a synthetic put option through a combination of cash and stock. The mechanics of the strategy demanded that the portfolio sell stocks as the market dropped, down to a floor below which the portfolio would not be allowed to go. At this point, the portfolio would be fully in cash. Conversely, the owner would switch from cash to stocks as the market rose. Portfolio insurance was born. Mr Leland developed the strategy with Berkeley colleague Mark Rubinstein and they went on with another partner to form Leland O'Brien Rubinstein Associates (LOR). Success bred success - and imitators - and by mid-1987 LOR's portfolio insurance system covered more than US$50 billion in institutional investors' assets, perhaps as much as US$70 billion. The fatal flaw of portfolio insurance was its need for a supply of ready buyers to soak up stock from insurance-driven sellers: in short, market liquidity. This condition was violated in the extreme conditions of the 1987 crash. At the time when insurance was most needed, it failed the test. It is a theme that was echoed a decade later by the near-collapse of the giant hedge fund Long-Term Capital Management. 'We never said there was a free lunch and we also warned all our investors that if we could not trade or if trading became too expensive, then our programmes would perforce be compromised and not deliver as they would in more normal kinds of market environments,' Mr Leland said. 'By 1987, because of our size, we were becoming somewhat concerned with our possible impact on the market.' LOR had a rough rule of thumb that if the market fell 1 per cent, the firm would have to sell about 2 per cent of client assets; by this time, this was effected through the selling of futures rather than the underlying stocks. It does not take a PhD in finance to realise that if too many investors were insuring their portfolios, such a dynamic could be disastrous. 'We were completely aware of this. We never said everyone can simultaneously insure their portfolios, because somebody has to take the other side of those trades,' said Mr Leland. 'We realised that we could stretch liquidity in the market.' He does not believe portfolio insurance caused the crash - 'we were merely reacting to movements in the market that already were taking place'. But he concedes: 'I think there is some truth to the fact that our trading did accelerate the speed of the decline.' He sidesteps questions over whether he felt responsible, but agrees it was a traumatic experience. 'Our offices were in Los Angeles. But given the large move on the preceding Friday, we anticipated we would be doing a lot of trading on Monday, and I actually flew down early Monday morning from San Francisco to Los Angeles. When I left, the market was down 60 points - that sounds like nothing today but you [have to] multiply by four - and I thought 'well, this is serious but not too serious'. 'By the time I got to Los Angeles the market was down 300 points and it was clear we were going to be doing a great deal of trading.' Mr Leland, still a professor at Berkeley, was in Hong Kong to deliver one of a series of lectures to mark Hong Kong University of Science and Technology's 10th anniversary, before going on to speak in Taipei and Beijing. He gave a presentation on his present research into optimal trading strategies, which shows how institutional investors can minimise transaction costs when rebalancing portfolios. Investors often have a target asset allocation: say 60 per cent equities and 40 per cent bonds. As market prices change, however, the asset mix changes, so investors typically 'rebalance' every quarter or year. According to Mr Leland's research, investors can save at least half their trading costs by establishing an optimal 'no-trade region', rebalancing only when the portfolio moves outside these bounds. However, his defining contribution to the annals of finance remains portfolio insurance. He stands by his ideas. 'Even though the dynamics of portfolio insurance have been criticised and I think unfairly . . . investors still really find such a product extremely attractive in terms of allowing them to invest with some degree of downside protection. People insure a lot of things and it makes reasonable sense to insure your assets as well as your house against fire. I think portfolio insurance will be a surviving product. The challenge is to provide it in a way that will be consistent with market liquidity and credit.' After the crash, portfolio insurance lost its allure and LOR lost much of its business. 'We were in some sense, not ruined, but hurt by our own success. It was the popularity of this process that eventually became too large for the liquidity of the market.' There were lessons from the crash. 'One is that pure competition when there is not general market knowledge of the overall size of positions can potentially be dangerous,' Mr Leland said. 'Nobody really knew how many portfolio insurers there were out there. And I think if that knowledge had been widely available, potential users would have been more cautious.' Most traders are driven by information: they have a view on the stocks they are buying or selling. By contrast, portfolio-insurance trading is 'informationless': it is purely mechanical, driven by a preset determination of acceptable losses. However, other market participants are not to know when selling is insurance-driven. Mr Leland points out that portfolio insurers sold only 0.2 per cent of total equity assets on October 19, 1987, yet prices fell by more than 20 per cent - far more than early financial-market models would predict. 'Generally the price you see does not affect your evaluation of a product. If the price is low you buy it and if the price is high you do not buy it. But financial markets are different, in that you may learn about others' expectations from the price. And in particular if suddenly a stock falls 20 per cent in one day, you do not run out and say 'hey, great sale, let's run out and buy the stock', you say, 'hey, what did somebody know that I didn't know?'.' Prices, in short, convey information. 'Our models suggest that if a relatively few people are much better informed than others, then the general populace, being sophisticated to know that other people know things they don't know, is very cautious in stepping in to buy, and therefore markets are a lot less liquid than they otherwise would have been.' The answer is more information about the size and nature of trading positions. 'If everybody knows perhaps that people are selling not for some information but for some hedging reason or something else, then they can say . . . it doesn't convey information.' He cites the example of Japanese telephone giant NTT, which on October 19, 1989, sold a dollar amount of stock which was twice what portfolio insurers sold in 1987 - yet the price of NTT stock barely moved. 'The reason it hardly budged is that investors knew in advance that this large sale was coming and they also knew it was not information-motivated. It was simply a decision by the government to reduce a nationalised company and put it in private hands. So investors knew it wasn't flawed goods.' LOR had recognised this potential problem and had even considered putting advertisements in the Wall Street Journal alerting people to the fact that if the market fell, it would have to do a considerable amount of selling for hedging purposes, in the hope this would dampen the impact of its trading. On reflection, it dropped the idea: 'We saw we could get into real trouble, maybe even with our own clients.' He argues it would be useful for a regulatory body to gather statistics on the potential amount of forced trading in the event of a market decline. Although difficult to achieve, this would enable investors to anticipate such events. Asked whether the risks in today's market were as big as they were in 1987, Mr Leland said: 'I just don't know. I doubt it, but nobody knows.' Graphic: hayn27gbz
