Lex sent me an interesting book review at the Economist of A Demon of Our Own Design: Markets, Hedge Funds, and the Perils of Financial Innovation by Richard Bookstaber. The gist of Bookstaber’s thesis is that, complexity of financial products, interconnectedness, along with liquidity and the resulting use of leverage, is bad for the financial markets:
Bright sparks like Mr Bookstaber ushered in a revolution that fuelled the boom in financial derivatives and Byzantine “structured products”. The problem, he argues, is that this wizardry has made markets more crisis-prone, not less so. It has done this in two ways: by increasing complexity, and by forging tighter links between various markets and securities, making them dangerously interdependent. As the system has grown more tangled, tougher regulation has only made things worse. Ironically, so too has the ocean of capital sloshing through markets in recent years. This has encouraged ever bigger bets with borrowed money, even though every seasoned investor knows that liquidity is the first thing to disappear when trouble strikes.
Worse, the fancy products cooked up by banks often have unintended consequences. That is because they are designed assuming rational behaviour, whereas markets have a nasty tendency to react unpredictably. Mr Bookstaber knows all about this, having played a key role in the creation of, among other innovations, “portfolio insurance” programmes. Instead of reducing risk and dampening shocks, as intended, these exacerbated the October 1987 crash.
I disagree with his thesis, and in fact, think the opposite is true. First off, financial crises tend to happen once since people vow never to let it happen again. The next time similar conditions appear, people prepare for it well in advance. So when the actual typhoon hits, the effect is smaller. For example people who see similar conditions of a crash brewing sell in advance, so when the actual crash hits, there are less sellers, so the effect is less.
Some reasons for the 1987 crash were program trading and “portfolio insurance”. Program trading is when computers execute orders directly based on programmed parameters. Portfolio insurance is when a fund is already long stock, and they think the market will go down, so they short index futures against their long stock. The perverse irony is that when the fund shorts the index futures, they are at the same time hurting their current long stock holdings. People see the futures go down, they sell their stock holdings, which begets more index shorting (or more portfolio insurance), which snowballs out of control. (A funny analogy about portfolio insurance I liked from before is as follows: portfolio insurance is like when your house catches fire, and you move your furniture 3 feet to keep it away from the flames. The fire moves closer and you move it 3 more feet away, etc, etc. ie it’s not really insurance.) Another reason for the ’87 crash was that everything was not electronic like it is now. People had to phone in big orders and such. So when the market went down, and program trading kicked in with the selling, people who wanted to buy couldn’t get their orders in fast enough – ie there wasn’t enough liquidity.
People making big bets also helps level out the market. For example people who short big against the market are natural buyers when the market goes down. If someone shorts ABC at 100, and it plunges to 50, he would probably want to lock in his gains and buy at 50. Heck, he can buy all the way up to 60 and still make out great. That’s why when a stock takes a plunge, many times you will see a bounce in the days following the plunge. If no one was short ABC, then there may not be someone looking to buy when it plunges to 50.
Interconnectedness and complexity of financial products are also net positives, in my opinion, simply because they provide more liquidity. When someone buys or sells something that affects something else, there is again, the natural buyer/seller at work – it broadens the investor base. For example, if someone wants to put on a complex futures trade, throwing in futures options along to hedge, that creates liquidity. The futures and options market makers will sell or buy from him, hedging their positions along the way. If the bet goes for or against the complex trade, there are natural buyers or sellers down the road. That was a bit of a convoluted example, but it’s similar to the short seller being the natural buyer above.
There is also now much more capital abroad than in the 1980’s. China, Japan, the Middle East, and Europe all have money, which translates to liquidity. If the US takes a hit, US investors may be gun shy, but foreign investors may not be. So they step in and buy. Net net, I disagree with him about reigning in leverage in general, because again, it provides liquidity. Long Term Capital Management (“LTCM”) was an extreme because their banks let them go beyond the normal guidelines. Perhaps if they didn’t let them go beyond the guidelines, the blowup may have been isolated. Here’s a good example: Amaranth lost billions because of a huge wrong way bet. However, this guy took the exact opposite bet as Amaranth and Brian Hunter, and made $2.4 billion, i.e., more than a third of Amaranth’s total loss went directly to his pocket. Not bad… Again, I think if anything, the benefits from liquidity, complexity, and interconnectedness outweigh the negatives. If anything they have a stabilizing effect on the financial markets rather than a destabilizing effect.
In complexity theory, the faster that a change in one part of the system propagates to the rest of the system the more stable that system is. Rapid change actually occurs when communication within the system breaks down. Freed from the dampening effects of the rest of the system, local areas deviate strongly until they provoke a local “avalanche” which then snowballs to the rest of the system.
Applied to financial systems, I suspect we would find that wild swings in the total market begin with the relative isolation of one submarket that prevents money and information from flowing in and out of the submarket. Lacking stabilization of the greater market, the submarket collapses trigging an avalanche.
Certainly, looking at the problem over the 500 years of capital market the pattern seems to show that the larger and more interconnected the markets have grown the more stable they have become. The US experienced 8 major depressions during the 1800’s, with numerous banking crises. The market used to seriously collapse about once every 15 years. Other countries and region show similar patterns. (Of course, its a chicken-and-egg problem, the markets are stable because they large but they could not have grown large unless they were stable.)
I think very few people have an intuitive feel for how complex feedback loops work. Most people do not see them operating in their everyday lives. We tend to focus on the dramatic avalanches that drive runaway downturns but ignore the prosaic positive feedback that drive the markets steadily upward day-after-day. This creates an intuitive sense in many that interconnectedness presents more dangers than isolation.