Hi Guys,
Try as we might, it is a challenge to be neutral about the marketplace’s direction. Everyone and his uncle has an opinion, a few more informed than most others.
That’s our forecasting curse, and it often does us more harm than good. We all like to participate in what often turns into a Loser’s Game. One extrapolation of the 80-20 rule can be used to establish the creditability of that argument.
You all are familiar with the generic 80-20 rule. One of its most popular interpretations is that 80% of the work is accomplished by 20% of the folks, or that 80% of an individual’s output is coupled to only 20% of his efforts. Lots of wasted motion.
The 80-20 variation that I want to discuss highlights the futility of forecasting follies. Trying to anticipate market movements rather than simply staying-the-course loses more often than it gains. The behavioral researchers have tested this hypothesis with an experimental game they call Red Light, Green Light.
Test participants in this game are asked to forecast if a random light will be either red or green. They are informed that the light color is randomly selected, but that it will be green 80% of the time. Now they play the game, and their score is recorded.
The expected correct score should be approximately 80%. Just about all experiment subjects fail to achieve that level. They fail because they believe they see a pattern that can be exploited. They are wrong. Here is a Link to a NY Times article that discusses some experimental results:
http://economix.blogs.nytimes.com/2011/02/17/forecasting-is-for-the-birds-and-rats/?_r=0A superior strategy, given that each outcome is randomly independent, is to forecast green every single time. From simple probability theory, if that strategy is used, the player will be on average (1.0 X 0.8) + (0.0 X 0.2) = 0.80 or 80% correct.
Most participants adopt a more complex strategy. Some like to play a strategy that is weighted to the given 80-20 distribution. Again, from a simple probability calculation, the player will on average generate (0.8 X 0.8) + (0.2 X 0.2) = 0.68 or 68% correct guesstimates. This strategy has decreased the odds of winning.
This simple probability analysis demonstrates the power and wisdom of betting on the favorite outcome consistently. Lower level animals learn this lesson quickly. Investors don’t. The strategy of betting in proportion to the frequency of occurrence lowers the likelihood of a successful outcome.
This type of analysis can be applied to the equity marketplace. The historical data reveals that on a monthly basis, equities have increased in value 59.6 % of the time since 1950. The upward reward happens roughly 70% of the time on an annual basis.
Doing the same analysis for a 70% positive annual equity return yields the following results for the two strategies defined. For the no-brainer who is always in the market strategy, that investor will obviously get positive rewards 70% of the time,
For the more sophisticated investor who plans to be committed to equities 70% of the time and in bonds/cash 30% of the time, his success ratio, on average, is likely to be (0.7 X 0.7) + (0.3 X 0.3) = 0.58 or 58%. Here again, the frequency strategy is likely to be less rewarding than the always-in strategy.
Unless an investor is prescient or especially insightful or perhaps just plain lucky, his forecasting (likely linked to a pattern seeking and seeing tendency) will probably degrade his cumulative long-term returns. Once again simple beats complex. Or does it?
There is a danger to oversimplifying a problem. I tend to be more or less committed to being in the market. But I do adjust my percentage of equity holdings depending on some measures like overall market P/E ratio. I hold fewer equities when that measure is high. As H.L. Mencken said: “For every complex problem there is an answer that is clear, simple, and wrong”.
Certainties do not exist in the investment universe. So I hedge and broadly diversify. What do you do?
Best Regards.
Comments
One may pick their turf here, depending upon what one monitors and/or feels comfortable.
Thanks for your interesting perspective. I agree that each investor gets to choose his own poison, although sometimes well meaning and well paid financial advisors intentionally usurp that responsibility. Regardless, investors are responsible for their decisions either made by themselves or some paid consultant.
I have had a dear and long-term relationship with one such advisor. I suspect we remain on friendly terms because I don’t invest through him, and I rarely accept his market pronouncements without careful due diligence.
The man is a flake. He is far too overconfident of his own knowledge base and the market’s projected direction. I suspect that he doesn’t keep score of his forecasts, but I informally do and its not pretty. He does play an aggressive tennis game, but, once again, he consistently overplays his perceived skill level. No surprise there.
It seems like every time one of his market forecasts goes haywire (quite often), he adds another parameter to his market prediction model. Not unexpectedly, the revised model reproduces the database with improved fidelity; that is, until the next market scoring period. If enough parameters are added, a near perfect match is likely.
Well, over several years, my friend’s market model has grown from about a 6 parameter model to perhaps a 15 parameter model. Yes, it does a better prediction today when contrasted to yesterday, but I suspect yet another revision will be implemented in the near future. If the number of open parameters are increased to exactly match the number of data points, a perfect reproduction (not a prediction) of the data set will happen.
Each of us has our own way of making a projection. As a general rule, the simpler the model, the more likely it will prove to be the better in terms of its robustness and longevity. Jack Bogle makes that precise point in many of his books.
But overall, precise market forecasting is a fool’s game. On an annual basis, returns are almost completely random in character. Adding parameters might make us more comfortable investors (a feeling of control), but they will not make us any wealthier than an Index-based plan. That might be a sad judgment, but it is fundamentally true except for a few very rare individuals.
I agree with Junkster. Everyone can control their money management with discipline, but forecasting the markets is an impossible task.
Best Wishes to All.
Often missing from critical discussions of forecasting is reference to the multiple methods and approaches used. Rather than attacking the general concept, it might be better if folks looked at the wide variety of techniques utilized. It is perhaps a disservice to those who dare voice their market timing schemes without addressing the specifics of how they attempt to do so. To that end I've linked a Wikipedia article addressing the topic in greater detail. At the end of my sermon I've attempted to list many of the forecasting techniques mentioned in that article. You'll find that several different techniques have been mentioned favorably by MFO participants over the years.
https://en.m.wikipedia.org/wiki/Forecasting
I agree that over very long time horizons, Bogle's reversion to the mean holds true. Since our life-spans and levels of patience don't always coincide nicely with those long time frames, I see some logic in investors attempting to forcast nearer-term movements - however imperfect it may seem. In effect: some are willing to forgoe some of Mr. Market's anticipated long-term returns for what they see as lower near-term volatity and a higher level of comfort.
Summary of forecasting methods - From Wikipedia (edited)
Qualitative method
Quantitate method
Average approach
Naïve approach
Drift method
Seasonal naïve approach
Time series methods
- Moving average
- Weighted moving average
- Kalman filtering
- Exponential smoothing
- Autoregressive moving average (ARMA)
- Autoregressive integrated moving average (ARIMA)
- Seasonal ARIMA or SARIMA
- Extrapolation
- Linear prediction
- Trend estimation
- Growth curve (statistics)
Regression Analysis
Autoregressive moving average
Judgmental methods
- Composite forecasts
- Cooke's method
- Delphi method
- Forecast by analogy
- Scenario building
- Statistical surveys
- Technology forecasting
Artificial intelligence methods
Artificial neural networks
Group method of data handling
Support vector machines
Data mining
Machine Learning
Pattern Recognition
Seasonality
Cyclical behavior
Point is: I think it a bit unfair to lump all forecasting approaches together under one umbrella in considering their validity or their desirability for various types of investors.
Thank you for your perspective and your extensive reference. The reference is overwhelming with detail that only a statistician would ever want to know.
Not all forecasts are equally formulated; not all forecasters are equally talented. Most forecasters have a dismal record when scored over time, but a few demonstrate a rare skill that exceeds a luck contributing factor.
We are very fortunate to have access to a dedicated researcher in this arena who documents his lifelong studies which are extensive in both time and manpower. That researcher is Philip Tetlock. His most recent book on the subject is titled “Superforecasting, the Art and Science of Prediction”.
The title broadly summarizes his fundamental conclusion: making a forecast includes elements that are both scientific and artful. His studies have generated much buzz and much controversy. He likes to classify forecasters into two groups: hedgehogs and foxes. The hedgehogs knowledge is deep in one specialized area; the foxes knowledge is more shallow and more diversified. Tetlock concludes that foxes are the better forecasters.
In repeated test periods, Tetlock has demonstrated small enhancements in prediction capability. For example, he has selected superior forecasters from earlier experiments and has formed them into teams. The teams show prediction improvements, perhaps because of the team interchanges. The improvements seem small, maybe just noise.
In his book’s closing statement, Tetlock says that “To be sure, in the big scheme of things, human foresight is puny, but it is nothing to sniff at when you live on the puny human scale”. That bit of philosophy aside, the documented improvements seem marginal, and yes even puny.
But some guys have proven to be better predictors than the generic crowd, just like Ted Williams was a superior baseball hitter and a two war ace fighter pilot. The reasons why are not so clear. Regardless of the forecasting skill set, we are typically more comfortable making or seeking forecasts. Perhaps in so dong we feel like we’re more in control.
The referenced Tetlock book may well exceed your interest and/or commitment time level, so here is a Link to a good review of that book:
http://www.economist.com/news/books-and-arts/21666098-forecasting-talent-luckily-it-can-be-learned-unclouded-vision
Once again, I see it much like Junkster does. Investing is part science and part art. The dominant part is art. There are many approaches to successful investing, each with their own advantages and pitfalls. It’s a good practice to be comfortable with your chosen method since there will surely be periodic ups and downs. An important element is to have “true grit” to persevere and remain committed through the turbulent turmoil.
Thanks again for your valuable participation.
Best Wishes.
As long as the freeways during rush hour and the corporate office spaces are bustling, then it is doubtful that this market cyclicallity will cease to be profitable ...
Regards,
Ted