Do Meteorologists Make Good Stock Traders?

By Chris Ebert

Try this: Look out of the window and, based on what is seen, make a prediction about tomorrow's weather.

If it's sunny today, does that mean it will be sunny tomorrow? Or, if it's cloudy today does that make it any more or less probable that it will rain tomorrow? That's just the sort of process that many participants in the stock market face on a daily basis. Traders can learn a lot about their own trading simply by considering what it is that makes a weatherman good (or not so good) at predicting the weather.

Perhaps, instead of just looking out of the window, a bit more information would be helpful – a few technical indicators, such as the temperature, the barometric pressure, and the wind direction. Even so, it is likely that the only thing those additional indicators would do is provide a vague forecast based on some very generalized probabilities. For example, hypothetically, 60% of the time when the barometer is falling and the wind is out of the northeast, rain is on the way the next day.

While a 60% chance of rain isn't likely to provide adequate guidance for someone planning an outdoor barbeque, a 60% probability is often a good enough probability for a trader. In fact, many traders learn to survive on 60% or less. But, what most traders would really like is to be 80% or 90% accurate – as accurate as the local meteorologist.

The trouble is, no matter how many technical indicators a trader uses, the accuracy will likely never approach such lofty levels on a consistent basis. Sure, there might be a few calls that are right on the money – anyone can make a stock prediction or two and be proven correct simply by chance – but no technical indicator (or combination or indicators) will be right 90% of the time over the long term, or anywhere near 90% for that matter.

The only reason the weatherman is able to be so accurate, so consistently, is because of a process known as modeling. Only when millions of bits of real-time data are collected and compared to millions of bits of historical data is the meteorologist able to make a consistently accurate prediction. As long as the real-time data has similarities to some set of historical circumstances that existed in the past, there is a high probability that the current circumstances will unfold in much the same manner this time as they did back then.

Historical data provides the basis for the model. The model can then be used to predict the future with a reasonable amount of certainty. It works for any process in which the future depends to some degree on the present. Tomorrow's weather depends upon today's weather, just as tomorrow's stock prices depend on today's.

It is entirely plausible that the stock market can be modeled just as easily as the weather, perhaps more easily. With weather models, it is necessary to capture data from some rather out-of-the-way places, such as weather balloons and satellites orbiting high above the Earth. Stock market models, on the other hand require nothing more than stock prices and prices of other financial instruments, all of which are available to almost anyone with an Internet connection.

The problem for an individual trader attempting to model the market isn't the lack of information; it's the sheer quantity of it. There is simply too much data for any one person to analyze.

Indeed, the atmosphere provides far too much data for any single meteorologist to analyze. The only reason weathermen have conquered their field to any degree is that their collaborative effort is directed at a single concerted purpose – modeling the Earth's atmosphere. While rivalry among peers is bound to exist in any field, weather forecasting does not have the same financial implications as forecasting the financial markets.

By and large, weathermen don't profit by withholding data or feeding others bad data; instead they succeed by building on the successes of their peers. Weathermen have thus very publicly succeeded in their field over the last 100 years while traders have, as a group, failed miserably. There is little reason to share successes in market-modeling with the public. In fact there are often reasons not to share. Unless there is a financial benefit, most traders are inclined to keep their successes to themselves.

An individual trader in the stock market today is about as qualified to make a prediction as a weather forecaster was in the 1800s. That said, forecasters in the nineteenth century learned to do the best with what they had to work with; and some of them were amazingly accurate given the limited amount of data available at the time.

Imagine if a weatherman from the 1800s had the tools commonly available to the public today. Would it help them make a more accurate forecast? Perhaps, but likely only because the data available today would tend to fill in gaps they already realized existed. For example, a single satellite photo of a hurricane would fill in a lot of missing information on storm development – an area in which a nineteenth century inhabitant would almost certainly be deficient, yet cognizant of his own deficiency.

To an individual trader in the stock market today, there are no satellite photos. There are no forecast models. That does not imply that such models do not exist. In fact, it would be naïve to presume that such models are not in existence. Big institutions certainly have the capacity to crunch the data – millions of bits of financial data – and produce their own models. But there is little reason to share the results unless it brings a financial benefit. They share the models only when it is convenient for them to do so, e.g. when sharing is likely to bring some financial benefit. The rest of the time, the individual trader is privy to only the raw data; any leaks of model data are likely inaccurate or intentionally misleading.

The individual retail trader is at a distinct disadvantage in the stock market today. Big institutions and large investors almost certainly rely upon highly accurate financial models to make their trading decisions – models as accurate as those that produce the local weather forecast. But, unlike the weather, the individual has no access to these financial models. The individual is forced to rely upon crude indicators to make financial decisions. It's like a 19^th century weatherman competing side-by-side with a 21^st century meteorologist.

Although such a predicament can seem depressing initially, a stock market trader does have hope. Trading today, for the individual, is like forecasting the weather without the use of sophisticated weather models. But, even though a trader must compete at a disadvantage, the raw data is available today more freely than at any time in the past. An individual retail trader generally has access to more raw data these days than could ever be analyzed.

Because an individual lacks the capacity to analyze all of the raw data now available, a realistic goal of a modern trader is to be like a 19^th century weatherman with 21^st century tools. The indicators available today are more numerous than they once were, and the delay in obtaining them has grown shorter. Indeed, many indicators that once took hours, days or weeks to obtain in print form are now available electronically, virtually instantaneously. One can only imagine what some of the legendary traders of history could have done with the data and indicators available in modern times.

An individual trader's best chance for survival today is not through modeling the financial markets to predict the future, but through utter and complete mastery of a few good indicators. The market has changed in the past hundred years or so. Technology has turned a once-level playing field into a game rigged by those who have the means to use technology to produce reliable models. There is just no way an individual can consistently compete with the accuracy of market predictions made by supercomputers, especially if the supercomputer's results are kept secret.

A modern trader can still make predictions based upon indicators, though, just as it's possible to forecast tomorrow's weather without a supercomputer. The trader will almost certainly be surprised from time to time, as the forecast won't always be accurate. But, in trading even a modest 60% accuracy is plenty, particularly if combined with a sound method of risk management such as simple stop-loss orders and reasonable position sizes.

There are some things that are fairly reliable in the stock market – indicators that have a good probability of predicting the near future without the need for modeling the entirety of the financial markets with a supercomputer. Bull markets tend to stay bullish, for example. Bear markets remain bearish. Trends tend to continue moving in the same direction. Volatility ebbs and flows.

Many traders eventually learn to adopt an indicator that suits their personal taste. Some, for example, become experts at spotting trends, others prefer to spot anomalies in the trend that portend a reversion to the mean. Still others look for cycles, such as those that commonly occur with volatility. Very few are experts of everything – they don't need to be – and it's unlikely they ever could be without an array of supercomputers and the manpower to run them.

Each week, the Thursday Evening Options Brief provides readers here with a snapshot of the stock market environment. Much like a weather forecast made by a backyard observer, the analysis is intended to give traders an edge. It is designed to allow traders in the stock market a unique perspective – an option trader's point of view.

While traders of stocks are concerned only with where stock prices are headed, traders of options are concerned also with how long it will take prices to get there. It's like looking at the humidity in addition to the temperature, to forecast the weather, or the rate of change in the barometric pressure instead of just the momentary value. Option performance gives a trader a view of the stock market environment much like a weather forecast.

As indicated in the Thursday Evening Option Brief, the stock market remains in Bull Market Stage 2. This is an environment known as the "digesting gains" stage, in which stock prices tend to rally higher only to experience periods of pullbacks or consolidation as the gains are digested by the market.

Until the options environment changes, the stock market will keep doing what it has been doing – digesting gains. Over the coming week or so, the only way the stock market weather will change is if the options themselves experience a change in profitability.

Bull Market Stage 1: If the S&P climbs above 2150 rather soon the increase in exuberance would add sustainability to the current rally.
Bull Market Stage 2: If the S&P remains in the 2100-2150 range through early March, the lack of enthusiasm could make long-term sustainability for the current rally iffy.
Bull Market Stage 3: If the S&P falls below the 2100 area in the near future the lack of momentum could doom the current rally to failure, at least for several weeks.
Bull Market Stage 4: Below approximately 2050 the S&P could be considered to be in a Bull-market correction.
Bull Market Stage 5: If the S&P experiences a major sell-off, such a decline is likely to be temporary as long as it bounces higher before reaching the 2000 level.
Bear Market Stage 5: Below S&P 2000 this March the environment would almost certainly be bearish, and the decline in stock prices would likely continue deeper.

Obviously the above forecast does not have the benefit of using supercomputers to generate models of future stock price behavior. But, the approach is a time-tested method for revealing changes in the stock market environment. It's like the forecast of a 19^th century weatherman using 21^st century tools.