Wall Street legend Peter Lynch once said: “In this business, if you’re good, you’re right six times out of ten. You’re never going to be right nine times out of ten.” What he’s trying to get at here is the unpredictability of the stock market. No matter what you think you know, a stock’s movement is uncertain at the end of the day. Hoping that you are correct (profiting from your trades) nine out of ten times is therefore futile. But this hasn’t stopped people from trying to do so anyway. Today, we are going to look at some attempts people have taken to do the impossible: predicting the stock market.
Fundamental Analysis and Warren Buffet
Yet another investing icon, only this time he’s the ninth-wealthiest person on Earth and Chairman of Berkshire Hathaway (the world’s most expensive stock, at a whopping $470,000 per share), Warren Buffet is famous for many things, one of which is his promotion of a particular investing style.
Value investing is buying stocks based on a thorough analysis of the company. Scrutinizing their balance sheet, their profit margins, and other financial metrics. What this approach is all about is determining if a company is undervalued. If it is, you buy the stock at less than its intrinsic value, what Benjamin Graham, the father of value investing, calls “margin of safety.” The thinking is that if you hold onto the stock long enough, the market will value the company at what it’s actually worth, thus earning you a profit.
Buffet, along with several other famous investors such as Charlie Munger (Vice Chairman of Berkshire Hathaway) and Seth Klarman (15th highest earning hedge fund manager in the world) are adamant proponents of value investing. But Buffett has his own spin to this tradition. The Buffet Indicator divides the market capitalization of all publicly traded stocks in the US (most investors use the Wilshire 5000 Total Market Index for this) by the current US GDP, thereby giving him a ratio of sorts to determine how the market is being valued. Generally speaking, if the ratio falls between:
50-75%, stocks are being undervalued
75-90%, appropriately valued
Above 90%, overvalued
But not all investors are convinced of its validity of course. A common criticism is that it completely ignores the profitability of businesses, with the only 2 variables being the stock market’s total market cap and US GDP. But even if you were able to get around this problem by replacing GDP with total US earnings, comparing the US equity market with earnings ignores the fact that American companies are exposed all around the world. Nearly 30% of S&P 500 revenue, for example, is international. Think companies like Apple, Amazon, and Coca-Cola - they’re all multinational firms.
Technical Analysis
In contrast to fundamental analysis is technical analysis, which concerns itself more so with a company’s volume and price data as opposed to their financial books and then, through behavioural economics and quantitative analysis, predict where the share price is going to go. Technical analysts employ far more complex indicators than a simple division operation like Buffet does. Markers such as relative strength index (measures the speed and change of price movements) and moving average (smoothens stock price data by constantly updating the average share price of a given company) are a day trader’s bread and butter.
While nearly all day traders use technical analysis, it should be noted that in no way are all technical analysts day traders. Investing of all time frames, ranging from minute-by-minute charts to graphs that span over year-long periods, employ technical analysis.
Elements of this approach first appeared as early in the 17th century with diamond merchant and financial expert Joseph de la Vega analyzing the Dutch financial markets, although Asia has a different story; they say that because 18th century Japanese rice merchant Homma Munehisa invented the candlestick technique, he in turn invented the entire methodology.
Regardless of who invented it, the practice is built off of a few key principles:
The efficient market hypothesis is true (more on what this is later)
Stock price movements exhibit broad, observable trends
History repeats itself, that is to say investors will behave in a similar fashion to investors before them
Unlike its fundamental analysis rival, there is a much larger debate as to whether technical analysis actually works. It is, however, difficult to empirically analyze its returns and compare them to fundamental analysis simply because methods within technical analysis vary greatly and so investors can sometimes come up with contradictory predictions regarding a share price’s movement.
Computer Models, Tobias Preis, and Jim Simons
Both technical and value investing can make way, for we are now in the digital era. Forget having to pour through a company’s internal documents or measuring their stock movement from the last 3 hours, why not just have a computer do it? Traders are now turning to artificial neural networks - algorithms designed to recognize relationships in massive data sets - to predict stock prices.
There are 2 pioneers worth highlighting here, one of whom put to work his academic research in investment endeavours. The first is Jim Simons - the mathematician who cracked Wall Street. Simons founded the Renaissance Technologies hedge fund in 1982, where he and his team of fellow mathematicians as well as physicists and statisticians put together automated programs wired to look for non-random movements in the stock market and then make predictions from there. It could be said that the process was just as tedious as it was intellectually demanding.
In order for his program to make meaningful predictions, he needed to force feed as much stock data as he could get his hands on. And that it wasn’t just a few USB sticks worth of info, no; Simons took and stored terabytes of data per day which were to be analyzed. All of this plus the fact that the data wasn’t available a few clicks away online.
But all this hard work paid off big time. His firm’s most profitable investment portfolio, the Medallion Fund, earned over $100 billion in trading profits since 1988, which means nearly 40% in average net returns. Simon’s net worth is now estimated to be $25.2 billion, making him the 57th richest person in the world as of writing.
The second key leader of this investment strategy is finance professor Tobias Preis, in which he found that online indicators can predict stock market moves. He specifically looked at search volume data for 98 terms of varying financial relevance provided by Google Trends and found that increases in search volume for these terms tend to precede large losses in financial markets. The three biggest terms in this regard were “debt”, “color” and “stocks”, in that order.
The Efficient Market Hypothesis and A Random Walk Down Wall Street
This controversial theory posits that asset prices reflect all information available to investors, which suggests that, when applied to the stock market, that a company’s share price is a product of all publicly known information. So if we assume the EMH to be true, that means changes in a stock price are either because of new information available or are simply random movements which fluctuate roughly around the company’s value based on the current information available.
Princeton economist Burton Malkiel published the implications of this in his classic A Random Walk Down Wall Street, where he popularized the random walk hypothesis, the theory that stock price movements are akin to that of a random walk. This is actually a rather self-explanatory mathematical term used to describe a path consisting of random steps on some mathematical space, such as the number line, shapes, or stock charts.
What this all suggests is that, given the randomness of the stock market, attempts to outperform market averages through stock market prediction are pipe dreams. As demotivating as that might seem, it cannot be stressed enough that the validity of both propositions are undergoing intense debate among economists and investors alike.
There still isn’t a definitive stock-market-predicting machine and there probably won’t be for at least a couple of decades. Then again, it’s possible someone already did it and cleverly decided to not share it with the world. Just saying…
Econ IRL
Earlier we touched upon the debate surrounding technical analysis’s success as an investing strategy. This week’s study is broadly considered to be an important one with respect to the aforementioned controversy. The authors found that, among a total of 95 modern studies, 56 find positive results, 20 obtain negative results, and 19 come up with mixed results regarding technical analysis’s success as a trading strategy.
However, the authors note various problems in many of the studies’ methodologies. Flaws such as data-snooping (where data sets are repeatedly analyzed with hopes of finding statistically significant results, even though there simply may not be any), difficulties in estimating risk when comparing investment strategies, and the ambiguity as to which strategies are considered “technical analysis” are ones that, the authors argue, stand in the way of putting together an accurate picture of technical analysis’s success.