The Futility of Predicting the Future

My father and I have a running joke on trading the stock market. In Chinese, the word “shares” is called 股票 (pronounced as “goo-pew” in Cantonese or “goo-piao” in Mandarin). The literal translation for each character is simply “stock ticket”. Phonetically, we joke that it sounds like “guess security” – that is, you have to guess the value of the security. Oh, how everyone wishes they had a functional crystal ball that allows one to figure out where the price of a stock is about to go. But alas, no such thing exists.

Thus far, the little experience I have accumulated has taught me that nothing predicts the market. You can’t use indicators because they are based off of past data. All they do is attempt to… wait for it… predict the future. But how can you predict the future when past data isn’t consistent? Yesterday, price moved one way. Today, price moved differently. No indicator could have predicted how price moves. Even in ranging markets, past data only goes so far until the ranging market ends and the indicator can no longer provide any semblance of what is about to happen next; the data is now corrupted by new, vastly different data from what it had used at the beginning of the range.

What about common patterns like triangles, double-tops, double-bottoms, flags, pennants, head-and-shoulders, etc.? To use these, you need a consistent way of drawing the patterns. That’s the first step. The second step is predicting future direction. Let’s say you have a consistent way of drawing double-tops. What happens after a double-top is formed? Will the market drop-off to oblivion or rally to new heights? Once again, we are stuck with the simple reality that no pattern can predict the market better than a theoretical coin flip. But that’s still based on the assumption of drawing the pattern consistently and correctly. I took to Twitter to look at how some people draw their patterns and all I can say is that everyone is a professional market artist. I have seen beautiful drawings full of vibrant colours and swirls. I have also seen the dullest and simplest of drawings. But I would surmise that a single black dot drawn on a white canvas would be worth thousands of dollars more than what you can win from simple, clean chart drawings. Everyone seems to see their favourite animal in the clouds. My conclusion is this: patterns don’t predict squat.

What about drawing support and resistance lines? Let me ask you this: what does price from 15 years ago have to do with the price of today? Everything has changed! The world economy was different then compared to now, the company has had to change to meet generational demands and new social values, and technology has changed the financial markets – we now have AI (or at least something very similar to it). Yet, everyone – even institutional traders – consistently repeat the same mantra: price has memory. But if you were to ask all these people to draw support and resistance lines on the same chart, you’d probably see nothing but swaths of black zones. No one draws the lines consistently. Even if we remove all the noise and focus solely on the most obvious support and resistance zones, you are still left with the most important question of all: where will price go next when it reaches a major zone? The correct answer is “flip the coin”.

Perhaps the key to successful trading doesn’t lie with such minute technicals. Perhaps, the key is to look at the company from a macro level and trade on a longer time frame. This is better known as The Warren Buffet Way. The Warren Buffet Way is simple: research the company and if it appears that it manages its financial affairs well, has a good product/service, knows how to market itself, then purchase some shares and hold onto it until retirement. But this method encounters the same pesky issue as the previous methods. Let’s assume you’ve found such a company. You buy the shares and hold onto the shares until retirement. At the time of purchase, how do you know that the company will still be around (let alone be profitable) by the time you retire? I had no idea that your research can predict not only the next day’s future, but two to three decades into the future! That’s absolutely stunning! And who would’ve thought that the majority of people who have had little business training would be able to properly analyse any random company and make such an informed analysis? I mean, I had to pay my dues in commercial insurance underwriting and financial risk management to learn how to do this but hey, according to Warren Buffet’s sage advice, anyone can do it! And so we are back to square one: the method of trading on fundamentals is no better than a coin flip.

I wrote in the primer that the only thing we can rely on is price. That is, if we can read and understand what price is telling us, then our trading will become successful. Even so, it’s wise to scrutinise this “technique” as I have with the other techniques above. The concept behind reading price is to understand market dynamics of supply and demand. That is, price moves up in order to discourage people from buying and it moves down to encourage buying. Price can also be viewed like an auction – the higher the price, the fewer people are wiling to pay for the shares. The lower the price, the more people are willing to pay for the shares. It is the median that’s “fair value”. Sometimes, when price is too low, almost everyone runs away. And the opposite is true – when it’s really high, most people perceive great value and jump on-board. Jesse Livermore made a similar observation – he said to go long when price makes a new high and to go short when price makes a new low. Just from these few points on reading price, I think it’s fair to say that there is no consistent way to read price! Even the theory can be defied with even lower lows and even higher highs.

I was reading blog articles and forum posts this week on reading price. Surely, there must be a technique to do this consistently. Alas, even those who say they make money regularly are relying on past data. To quote one writer who wrote a lengthy primer on trading based on reading price:

Here’s where data collection comes in, a means of learning the market’s language (what the trader thinks, feels, or believes is not a factor). And if the market has suggested in the past — over many series of trades — that 36% is no big deal, particularly as price remains well away from the next level of interest (the high of the 0300 thrust), then the trader may elect to stay with it. On the other hand, if the market has in the past suggested that the trend is in real trouble here, the trader, if he’s trading multiple shares or contracts, may redeem some of them at this point. Or exit entirely. It’s the market’s way of separating those who’ve done the work from those who haven’t.

Notice what he says: you must rely on past data, past experience, to form a directional bias. This is effectively prediction. In other words, despite his best intentions (and I’m sure he makes money), the method is as effective as a coin flip. You still can’t successfully predict the direction of price even with a vast understanding of market auction theory. Even the argument that past experience helps narrow down the choice doesn’t make sense. You have only two choices at the point of entry: price will move up or it will move down1Price cannot range sideways at the point of entry because your order will affect the price in some way.. That is not to say that learning M.A.T. is of no value. But don’t be deceived into thinking that understanding the ebb and flow of supply and demand will improve the odds of success. Mathematically speaking, knowing past behaviour does not reduce the 50/50 probability at the point of entry. It is still a coin flip at the end of the day.

So what can be done if at the point of entry, no matter what we do, the probability of winning is still 50%? At this point in time, I can only think of one action: financial risk management. You can’t improve the odds that price will go the way you want it to but I think you can reduce the impact of price going in the undesired direction. In insurance underwriting, we take a similar approach; you can’t predict whether the account will incur a claim but you can add deductibles, reduce limit exposure, add policy exclusions, and charge higher premiums to offset potential losses. In trading terms, this could manifest in reducing the size of your trades (that is, trade small). If your trades are small, the financial impact to your account will likewise be small compared to the impact of a loss to a larger trade size. In addition, if your trade is obviously a losing trade, then have the discipline to dump it. Let’s be honest, if you’re “hoping” for a reversal, it’s probably time to let go of it. Like Elsa would say, “Let it go, let it go.” To be honest, I think these are really the only two sure things one can do to reduce risk. Others have suggested trading with the trend (“follow the herd”) but it is difficult to pick up on the hints and signals that a trend is about to end. Even then, it is based on past experience which is not necessarily an indication of future behaviour. Others say to ensure you have appropriate stop losses. However, this begs the question of how you set those levels in the first place which, after some thought, appears to be just as arbitrary as an entry. Regardless of what you do, the key then isn’t in predicting the future or attempting to change the 50/50 probability but rather, the key is managing financial risk.

Scroll to top