What Is This?

There's a picture hanging in my office. Actually, a series of pictures published in Life Magazine back in 1940. The understated headline describes the sequence simply as "Tacoma Narrows Bridge Crashes in Puget Sound." The three photos show the bridge twisting, then breaking, and finally standing with its mid-section entirely gone. These pictures and other related information are widely available and continue to be studied and noted by civil engineers.

My primary interests are the worlds of technology and finance, and particularly the intersection of the two. In the photos of the bridge, I find a reminder of our human fallibility, and a catastrophic illustration of what can happen in any field of endeavor when we push the limits of our knowledge or seek to apply it in new untested ways. The bridge did not fall because its engineers were incompetent or because the construction company cut corners. Likewise no unanticipated natural or man-made disaster caused it to fall.

The bridge fell because it was subjected to forces that the engineering science of the time did not consider important. In fact, they did not really consider them at all. The existence of these forces was not unknown, but their impact was discounted because no structures had ever been built for which those forces became significant. As materials and engineering advanced, bridges became longer, and forces that had been ignored as irrelevant increased. In Tacoma in 1940, those forces passed an unrecognized threshold and announced their importance catastrophically on the morning of November 7, 1940, when strong winds demonstrated conclusively that bridge aerodynamics could no longer be ignored.

Interesting, but what does this have to do with anything?

Fast forward to October 19, 1987 and across the country to Manhattan, where a large group of people spent much of the day gathered around a Quotron terminal, watching the U.S. equity markets implode. Among the people gathered there were the technical geniuses who had devised Morgan Stanley’s program-trading capabilities. Throughout that morning, each minor upwards blip in the averages resulted in the same outbursts: “That’s it, the programs are kicking in, it’s all over.” Of course all those voices were wrong, it wasn’t over. By the end of the day the Dow Jones Industrial average had suffered its largest ever decline and the health of the financial system was seriously threatened.

What those people didn’t know, in fact what they couldn’t know at the time, was that on that morning in 1987, the framework on which their programs were based – a concept called “portfolio insurance” – had failed catastrophically. Not because it was wrong, but because like the Tacoma Narrows Bridge it included unstated assumptions and limitations that were not well understood. Most specifically, portfolio insurance depended on the existence of an efficient market. Market efficiency in turn depended on a regular, orderly, liquid market. When everybody ran for the exits on the morning of October 19^th, liquidity disappeared, and with it efficiency. The unstated foundation on which portfolio insurance had stood was shattered, and with it the markets and fortunes it was designed to protect.

Less than a decade later, FoxMeyer Health Corp was forced into bankruptcy after adoption of new order processing technology crippled the company. The experts of the software world had assured company management that the new solution would work, and in fact the SAP software selected was already acknowledged as a technology leader for the software market. The market for FoxMeyer’s products was growing strongly and appeared to demand new capabilities. The project was unveiled in mid-1994.

By the time FoxMeyer declared bankruptcy and sold its primary operating subsidiary to a competitor in 1996, it had become clear that all the unstated key assumptions were wrong. The business growth that management had anticipated never materialized when one of their key customers filed for bankruptcy. More importantly, the software never worked as advertised. FoxMeyer discovered – catastrophically – that while the software they had purchased had been used successfully in many organizations larger than theirs, it had never been used by a large distribution company and that it failed under the stress of hundreds of thousands of orders. The consultants, the software provider, and many others were blamed for the collapse.

Are you some kind of disaster freak?
Why talk about disasters?

I talk about them because I believe strongly that in any field of endeavor, the first rule of success must be “don’t fail.” While this is apparently obvious when stated so simply, it is a rule that’s often ignored. Investment managers talk about “winning big,” then put their clients’ money into investments that involve significant – but often unstated or ignored – risk of a major failure. Technology managers develop, buy or integrate fantastic solutions, but often don’t bother to understand the underlying limitations of the technologies they’re working with or the markets they are selling them to. Examples are plentiful.

Most of these losses and failures can be avoided by simply digging a bit deeper than most people (heretofore referred to as The Crowd) are willing to. Often it is not necessary to dig very deep, as the signs are right there on the surface for anybody willing to pay attention to them. I was there the day the market crashed in 1987. It didn’t come from nowhere. Signs of unusual problems were everywhere. The dollar was sinking, many individual companies were in trouble and trading was increasingly volatile. The Crowd ignored these signs because the experts assured them that a large-scale collapse was impossible. They were wrong.

I was, of course, not there when the Tacoma Narrows Bridge collapsed, but the history is clear. The bridge had been nicknamed “Galloping Gertie” by locals because it bucked and swayed violently in the wind. Again, The Crowd decided it was OK because the experts assured them that all was well. The unusual behavior was simply to be expected in so long a span. After all, the best of modern engineering had gone into the bridge’s design. The Crowd and the experts they depended on were proven wrong.

I was also not there in person when FoxMeyer made their own series of catastrophic decisions, but I was in the business, then just entering the fringes of what would become a powerful wave of ERP software development led by SAP. The Crowd, and clearly most of the insiders at FoxMeyer were 100% behind the early building of this wave. But it was clear to those of us involved in enterprise software in 1994 that any large business “betting the farm” on the new software needed to do so with the utmost of caution because the earliest versions of these products were poorly understood. Following The Crowd became deadly for FoxMeyer, and would continue to be hazardous to many companies for years to come. Several years later I was there when Hewlett Packard made the first $19m mistake in a series of software-related miscalculations that ultimately cost over $600m.

I believe that people willing to go Below The Crowd, to make observations of what’s really going on, can significantly reduce their chances of getting caught up in a financial or technological disaster. They may also significantly increase their odds of finding an overlooked gem that The Crowd of experts and their followers have ignored. Often this approach requires being a bit philosophical or a bit of a futurist. There's nothing wrong with that, so long as you remember that musing about the future is OK, counting on a desired outcome is an invitation to disaster.

Epilogue, or “What did The Crowd learn?”

To this day, virtually all asset pricing models used on Wall Street and elsewhere make the unstated assumption of a liquid, efficient market. The LTCM crisis in late 1998 was predicated by the failure of that assumption. Here and there we find interesting new ideas, but these are out of the mainstream and not widely followed.

The world of civil engineering has improved significantly since the 1930s, when the Tacoma Narrows Bridge was designed. Computer modeling allows the simulation of all sorts of forces that could not have been easily tested at that time. This, combined with the use of new materials, allows us to once again build elegant, minimalist structures. However, even here, we often don’t know what we don’t know. The failure of composite tail surfaces on two Airbus aircraft in recent years are beginning to suggest that our understanding of the materials many of these structures depend on may be lacking, even as The Crowd of engineers, designers and consumers continue to push for their greater use. And events of September 11, 2001 remind us that even the best engineered structures are only as capable as their designers imagined they needed to be.

Technology remains a minefield. Caught between the threat of technology obsolescence and the possibility of new technology failure, managers, engineers and consultants continue to often make poor and ill-advised choices often choosing to follow The Crowd of their peers rather than to look below it and learn their own lessons.

In short, The Crowd’s learning has been slow. If this website helps accellerate that learning in any way, then it will have served a useful function.