The Snow Storm That Wasn’t
Last Tuesday night New York city simply shut down and became like the movie “I am Legend”. At around 2:30 in the morning that day when I shuffled downstairs to get my coffee from the lone 24 hour a day bakery open on my block, Broadway was literally barren. No cars. No people.
The cause of all this panic was the “Nor’easter of the century”, which was supposed to drop two feet of snow on Manhattan and paralyze life as we know it.
Except that it didn’t.
The storm came, just as meteorologists predicted that it would, but its core missed the city by 10 miles. Eastern Long Island got buried. New England got buried. But by 5 AM the city sanitation trucks were basically plowing gravel in Manhattan and soon thereafter, embarrassed officials began to reopen all the city services that they shut down.
Some people say that weathermen exist to make economists look good. I think that’s patently unfair. Weathermen are much better at their job than market forecasters, but the events of the past week illustrated the limits of their profession and in turn provided some very valuable lessons for us as traders.
The National Weather service forecasting model relies on very a sophisticated statistical analysis process that makes millions of calculations per second to deliver its conclusions. Yet with all this brain power it failed.
Because statistics, contrary to popular belief, are not some ironclad scientific proof of truth, but rather a hodge podge of assumptions that can deliver startlingly wrong conclusions when even one tiny factor of the model is incorrect. That’s exactly what happened on Tuesday. The model was generally right but specifically wrong as it miscalculated the path of the storm.
Think about that the next time someone tries to impress you with a complex trading strategy with a thousand indicators. I can almost guarantee you that whatever marvelous ten million sample backtest they show you will most certainly lose money the moment it starts trading in real life. This will happen not because the algo writers are charlatans -- they most likely are very earnest, hard working people who will be bewildered just like you as to why their pot of gold has suddenly turned into a pile of sh-t.
The reason of course is that market variability will make mincemeat out of the most complex well thought out strategies because tomorrow is never quite like yesterday. That why ironically enough, the most robust strategies in the market are so simple that a child can understand them.