Soros et al vs Buffett et al

[locutusoftexas]

The most recent thread pays homage to George Soros and Jim Rogers, who have publicly declared the doom of our society and financial system.  All comments on CNBC range from caution to anger to personal capitulation.  Any amateur nontrader who has been paying attention must now have decided to keep holding whatever remaining stocks he or she has after selling in response to the extreme negativity that we have seen since August.

 

Warren Buffett, Bruce Berkowitz, and similar investors are buying.  They and their investors have all lost money in the last few months. Short-term paper losses happen to value investors, even the best; this is nothing new. I distinguish these knowledgeable fundamental investors from gunslingers like Bill Miller, who has finally been caught with his pants down.

 

I cannot tell you how many times over the last 20+ years I have read or been told that Buffett is lucky and his luck has run out "this time."  There is no doubt that Soros is a tough and resourceful guy and further, no doubt that he is a true trader, able to change his mind and his investments on a moment's notice.  He also likes the spotlight and he is presently selling his latest book.

 

However, even George Soros cannot tell the future. Yogi Berra, who proves much smarter with regard to this topic, tells us that "It's tough to make predictions, especially about the future." This is obvious, since the economies of the world have too many degrees of freedom for even the ~100 billion neurons in George Soros's phenomenal brain to follow.  George is overstepping his bounds and in fact the bounds of our physical universe. Periphrasing Bob Dylan in an interview during the 1960s, my or your predictions are every bit as good as George's at this point.  George's predictions are so much hot air, and, as a trader, he will forget them as soon as they prove false or true.  This is like speculating on a coin toss and therefore provides no particular insights to us.

 

Let us return to Warren, who has commented and has been quoted often about the short and long term views of markets.  Rather than speculate on today or next month, how about considering the risk-reward of the market over the next several years.  Based on history, dating back to the early 1800s, one can give bounds to the range of prices over the next 5 years of up 100%+ or down 60%.  These are bounds on the range of prices, not predictions of the actual range from this day forward over the next several years. The upper bound of up 100% (or more) is based primarily on the action of the bear market of 1967-82, which the present market pattern resembles very closely and which bears many economic similarities to the present. The  negative lower bound is based on the result of the contraction of the money supply by 1/3 during 1929-32 and therefore ignores the fact that the current US administration is not going to allow a similar contraction in money supply to take place.

 

Personally, I believe that neither Warren nor George Soros has been "lucky." However, I further believe that Warren understands stock valuation and stays within his circle of competence. With regard to actual investments Soros probably does, too; however, his public pontification takes him far outside of his or anyone else's circle of competence. With regard to predicting the downfall of our society or economy, I think that he should invest in a copy of Yogi Berra's quotes and stop embarrassing himself.

 

Meanwhile I thank Soros for the contrarian signal and attach a much higher probability to the above upper bound than the negative lower bound.

 

This board is populated by really smart and astute investors.  I am sure that they can provide a much better set of upper and lower bounds to medium term future market prices than I have. I look forward to hearing your estimates and reasoning.

 

Regards,

Tex

[scorpioncapital]

It would seem that Warren's point of view is that anybody who opens their mouth about macroeconomics is playing the wrong game. They are playing a game, but it is not the 'investment in a business' game. Unfortunately, while waiting, people don't have anything else to talk about. But if the filler material impacts your investments or emotions, that is a bad thing.

[ubuy2wron]

I believe the major difference between Soros and Buffett is Soros  thinks he can predict the future and Warren knows he can not.

[locutusoftexas]

My original post specifically said that neither Warren nor George is lucky.  You may also note that Soros has been predicting the disintegration of our economic system. So in fact he does believe that he can predict the future; otherwise he would state his prognostications differently.  He has ascribed his own success to being able to recognize when his predictions are wrong.  This is the part of the equation that is not often cited.

 

I believe that Soros is a great trader and a hard worker.  I also believe that, with his book and predictions, he has obviously stepped outside his circle of competence.

 

Almost all predictions are no more accurate than a coin toss and are not worth publicizing.  He might be right in his current grandiose predictions; that would be luck.  Kondratiev went to prison for demonstrating that historically, the US economy cycled between boom and bust but survived intact over the course of such cycles.  I will bet with history on this one, but as Dirty Harry said, I know my limits.

 

Regards,

Tex

[TerryO]

The procedure by which a prediction is made is called a "model." My job is to build models. I'd like to share a bit of what I know in the context of the Soros vs Buffett thread.

 

Over many generations, university statistics departments, economics departments and business schools have instilled the belief called "frequentism" in their students. Frequentism is the belief that the probability a real object will be found in a given state of nature is the limiting relative frequency of this state. The "limiting relative frequency" of a state is the proportion of occurrences of this state in the limit of an infinite number of observations.

 

According to people who have studied the matter, most practicing scientists subscribe to this belief; as frequentism was drilled into them by their professors, this is no surprise. Frequentism, however, is a trap, for it neglects the phenonenon which communications engineers call "noise." Because of this neglect, when a model is built under frequentism, it exhibits the phenomenon which I call "regression to the base-rates." This is that, when the model is tested, the observed relative frequencies lie closer to the base rates than the predicted probabilities. The prediction contains less information than the frequentist believes it to contain and may contain no information at all.

 

Building a model that reliably predicts outcomes in financial markets is difficult and sometimes impossible. In lieu of a reliable model, one strategy is to gamble on regression to the base-rates. This, roughly speaking, is the strategy called "value investing." Because of the pervasiveness of frequentism in peoples' thinking, value investing has been a fruitful strategy.

 

Investors exhibit belief in frequentism when they confuse signal with noise. For example, there is a 1 quarter drop in earnings and they take this to be a signal to sell. The value investor takes this "signal" to be noise and (if the price is right) buys.

 

Judged by their actions, Buffett gambles on the reliability of regression to the base-rates. Soros gambles on the reliability of predictive models. That Soros has been so successful suggests he has somehow evaded the trap of frequentism, thus being able to distinguish signal from noise.

[locutusoftexas]

Terry,

 

Thank you for a most interesting post. I need to read it carefully and think hard before I say much.  However, I will say this: For around 10 years (off and on), I have worked on understanding space weather forecasting in the context of meteorological forecasting.  Our best shot at making predictions are systems like this, for which we believe that we actually know the dynamical equations and processes underlying the phenomena to be forecast. To summarize: even in such cases forecasts are very untrustworthy, because our underlying observation set is relatively sparse for any given preceding epoch (say 6 hours, which is used for some types of meteorological forecasts).  Regarding forecasts of economic time series, I am virtually certain of the present impossibility and in fact the future impossibility (a prediction itself, which one must distrust -- Ha Ha. Also, "impossible" is a bad word to use but I use it just to express the idea of extreme difficulty.) In this latter case, we do not have accurate dynamical equations governing the system, which is effectively infinite dimensional (in terms of degrees of freedom) and further involves human psychology, which is very poorly understood.

 

Among the topics that we could discuss in this context are persistence and multivariate regression-based forecasting, as well as data assimilation.

 

I believe that Soros is quite disciplined and his speculation category might place him with Jesse Livermore. The risk reduction strategy involves placing severe constraints on the amount of time that is allowed to pass in order for the expected scenario to take place. In contrast, my impression is that Graham and Dodd and, by extension, Buffet could actually estimate with superior accuracy the actual underlying and realizable value of specific equities and similar investments. The latter investors are willing to allow sufficient time (i.e., a long time) to pass for that value to be recognized in the marketplace. That is their risk reduction strategy. Both investor types require exceptional discipline to be highly successful and I expect that neither is reducible to equations or algorithms.  That is just my opinion, somewhat based on the severe limitations of our ability to predict the economy or the markets based on merely quantitative data, which themselves are almost certainly too sparse to describe even a past state accurately.

 

Thanks again. I will study your post but may have nothing further to add.

 

Tex

 

 

[SharperDingaan]

 

You might want to keep in mind that these are two masters at their own facet of investing, & that their processes both rely on human frailties in order to work.

 

WEB essentially waits for the price to fall below, & then rise above, his estimate of IV net of MOS. A static approach that uses time as its major variable. Soros essentially looks for inductive logic fallacies & takes the other side of the trade. A dynamic approach that uses human behaviour as its major variable. Both use 'Mr Market'. WEB takes him up on his 'irrational' offers, Soros takes the other side of his trade.

 

Then consider that Graham was in many ways a combination of the two. He essentially challenged the accepted convention (Soros?) & came up with a different approach that has since been improved on (WEB?)

 

Who's better? is not really a fair question.

 

SD

 

 

 

[TerryO]

Tex

 

I believe you'd be interested in what I'm doing. There is an overview at http://www.knowledgetothemax.com.

 

One of the applications of the technology I'm trying to peddle has been to meteorology. This application was astoundingly successful in the sense of increasing the time span over which precipitation could be forecasted by a factor of 12 to 36. The original idea was to amalgamate predictions from a mechanistic model based upon solution of the equations of motion of the atmosphere with predictions from an empirical model based upon observational data. However, non-linearities in the equations of motion made the mechanistic model unstable over time spans of a few weeks, the so-called "butterfly effect." Thus, the mechanistic model was abandoned for a purely empirical model. In this approach, an optimal decoder was built to decode "messages" from the future, where the "message" was the sequence of precipitation outcomes. Taking this approach increased the span of time over which precipitation could be forecasted from 1 month to 12 to 36 months. As it turned out, there is an oscillation in the Pacific Ocean called the El Nino oscillation with a period of a few years. By tuning a decoder to this oscillation, one can predict precipitation in California several years in advance.

 

A similar approach has been successfully applied to prediction of financial time series.

 

Terry