Predictions: Popular, but not very helpful

Market participants trade expectations about the future on financial markets day in and day out. So, it’s hardly surprising that an entire profession devotes itself to concocting forecasts about stock prices and economic variables and that those projections, in turn, are daily grist for the financial media. Even retail investors like to consult the (point) predictions made by purported experts. But all forecasts should be treated with caution and shouldn’t distract investors from sticking to a long-term investment strategy.

Let’s flip a coin

Who wouldn’t want to know what will happen tomorrow? The ability to see into the future could be put to use very lucratively, particularly on financial markets. A look into the crystal ball is ventured anew every day on stock-trading TV shows when so-called “gurus” draw attention to themselves with their market views. One can safely question the added value of their predictions, and that’s exactly what a number of studies have already done.

An investigation by CXO Advisory Group examined the extent to which self-proclaimed “experts” (or those endorsed by others as being experts) are useful guides to timing the stock market. To find out, CXO Advisory Group collected and analyzed a total of 6,584 forecasts from 68 experts on the US stock market over a period from 2005 through 2012. The experts’ forecasting methodologies covered the entire spectrum of conceivable approaches, ranging from analyzing technical indicators to fundamental analysis and observing market sentiment. The investigation found that the accuracy of the forecasts was below 47% on average, which is worse than simply flipping a coin. The distribution of forecasting accuracy almost exactly matched a typical bell curve, just like one would expect to see from a purely random outcome. The highest individual accuracy rate among the experts examined was in the area of around 68% and the lowest was in the vicinity of 22%. David Bailey et al. came to a similar finding in their study titled “Do Financial Gurus Produce Reliable Forecasts?”, which explicitly references the S&P 500 index. In that study, only 48% of all forecasts were correct, and two-thirds of the gurus were wrong in the majority of cases. The roster of “participants” in both investigations included well-known names like Jeremy Grantham (the founder of the value-oriented investment house GMO), Marc Faber (the author of the “Gloom, Boom & Doom Report”), and Jim Cramer (the CNBC superstar). Cramer ranked among the worst competitors, with a forecasting accuracy score of 47% in the CXO study and 37% in the study by Bailey et al. However, Cramer’s low accuracy rate thus far hasn’t harmed the success of his stock-market show “Mad Money”, which has been running for almost two decades now. The reliable failure of his market predictions to come true actually even culminated last year in the launch of an Inverse Cramer Tracker ETF. Meanwhile, the Long Cramer Tracker ETF that was launched at the same time was shut down just a few short months later due to a lack of investor interest.

Upshot: Stock-market tips from gurus don’t cost you anything, but they don’t provide any added value and may actually even have negative value if they distract investors from sticking to well-thought-out strategies. This risk particularly looms when predictions by gurus merely confirm an investor’s own personal opinions (confirmation bias) and undermine objectivity. Assertions by purported experts should be viewed as being no more and no less than what they really are: purely entertainment.

No better than flipping a coin | Even gurus don’t have a crystal ball

Guru accuracy histogram

Sources: CXO Advisory Group, Kaiser Partner Privatbank

Crowd wisdom is no help either

If it is not profitable to trade on the instructions of individual gurus, maybe it helps to consult the wisdom of the masses, i.e. the “predictions” and expectations that can be deduced from surveys. Investigations concerning this as well have already been undertaken, for example in the study titled “How Accurate Are Survey Forecasts on the Market?” published in March 2023 by Songrun He et al. (Washington University). The researchers analyzed three established surveys in which respondents are polled on array of questions including their outlook for the US stock market for the next 12 months. The three surveys they examined were:

• The Livingston Survey, a semiannual survey on the US economy (and the S&P 500 index, among other questions) conducted by the Federal Reserve Bank of Philadelphia (in June and December). The survey polls experts and economists from the financial industry and academic institutions. Around 90 participants are on the survey’s mailing list, and 55 to 65 of them respond to the survey each time.
• The CFO Survey, a quarterly survey of chief financial officers in the USA inquiring about the financial outlook facing their respective companies. The survey polls a sample of around 4,500 participants (of which around 400 respond each time).
• The set of data compiled and analyzed by Stefan Nagel (University of Chicago) and Thengyang Xu (City University of Hong Kong), which consolidates various data sources – including the UBS/Gallup survey, the Conference Board survey, and the University of Michigan Survey of Consumers – to form a representative survey of a typical US household.

Three surveys and three groups of respondents with differing expertise on financial matters. What could possibly go wrong? A whole lot, it turns out, as illustrated by a simple statistic: none of the surveys came close to correctly predicting the actual (historical) stock-market return. The actual long-term return (since 1945) amounts to +7.5% per annum on average, but the average annual return projected by all three surveys was only half as high at +3.8%. But not only were the absolute return estimates far detached from the reality of the market, so were the predictions of the market’s direction (up/down). Neither financial-market pros nor consumers were better than a naive prediction simply projecting a repeat of the previous year’s return. Only CFOs were able to deliver a small, but statistically insignificant, degree of added value with regard to the market trend.

Upshot: The crowd is also inept. There is no such thing as crowd wisdom, at least not with regard to (stock) market predictions. Whoever can’t get around having to quote an estimate would do best to simply assume a random walk by the market.

Year-end forecasts

Even though the evidence of the scarce utility of predictions is more than abundant by now, forecasting continues to be a permanent fixture in the world of finance. Highly specific point predictions, for example, are part of the daily bread for single-stock analysts as well as market strategists. Crystal ball gazing regularly reaches its climax at the turning of the year in the form of new year-end targets for blue-chip stock indices. Analyzing the year-end forecasts from the last 20 years for the USA’s S&P 500 index, one discerns a certain pattern, to wit: market strategists who make top-down forecasts of the year-end closing level of the benchmark US index by factoring in the macroeconomic outlook are notoriously pessimistic and have projected an average annual return of just +3.4% since 2005, which is far below the actual realized return of +9.0%. In contrast, single-stock analysts, whose point predictions on individual stocks get aggregated into an implicit expected index level (using a bottom-up approach), are systematically overoptimistic and projected an average annual return of +11.8%, which was well above the actual realized performance.

Overly optimistic/pessimistic | The historical return is the better prediction

Projected and actual returns for the S&P 500 index

Sources: Bloomberg, Kaiser Partner Privatbank

A look at the index forecasts for 2024 reveals the same picture of pessimism and optimism. Market strategists see stagnation, whereas single-stock analysts anticipate double-digit percent share-price gains on aggregate. What are investors to make of this? They shouldn’t take the projected numbers at face value, or else confusion is guaranteed. If a person had simply given a prediction tip in line with the average annual return on the S&P 500 since 1945 (+7.5%) for each of the last 20 years, that person would have been a bit too negative, but would have put the community of experts to shame in the matter of forecasting accuracy.

Confusion squared | Market strategists and stock analysts are in disagreement

Index forecasts for 2024

Sources: Bloomberg, Kaiser Partner Privatbank

Upshot: Investors shouldn’t care about share-price and index targets because even the experts who concern themselves with this subject every day haven’t the slightest idea where prices will be at the end of the year. However, their work isn’t entirely futile because stock and market analyses at least identify the relevant price drivers and can thus help investors to recognize risks and exploit opportunities.

Central-bank officials also don’t have a crystal ball

If anyone at all is capable of making good predictions, then it has to be central banks. One could at least make this assumption regarding their interest-rate forecasts. After all, not only are central-bank officials “genuine” experts who have access to vast quantities of economic data, but they also set policy interest rates themselves. But even that good point of departure provides no protection against misjudgments, as the forecasting performance of the last three years strikingly illustrates. Back in late 2021, the US Federal Reserve expected that it would raise its policy rate to a maximum of 2.1% by the end of 2024. In reality, though, it actually had to hike the federal funds rate to over 5% by as soon as summer 2023. The erroneous interest-rate forecast was caused by another misjudgment: the inflation forecast. At the end of 2021, the Fed expected inflation to recede from 5.2% to 2.6% over the next 12 months. In reality, inflation was still hovering at 4.9% at the end of 2022. The past pandemic and post-pandemic years admittedly were extremely difficult to assess, let alone predict, arguably for everyone. But even in less volatile times, estimates by central banks have tended to exhibit a rather short half-life in the past.

Disillusioning | Even central banks have difficulty making forecasts

Policy interest rate and inflation in the USA (and projections by the Fed)

Sources: Bloomberg, Fed, Kaiser Partner Privatbank

Upshot: The future is uncertain. That’s why markets and macro variables are impossible to predict with pinpoint precision. Even sophisticated, elaborate models are of little help here. For this reason, the experts at Kaiser Partner Privatbank deliberately abstain from making point predictions on markets and macro variables. In the Monthly Market Monitor, we confine ourselves to keeping track of consensus estimates on economic growth, inflation, and interest rates, and we only make forecasts on interest-rate trends. To those who may think this practice seems to be lacking in ambition, we highly recommend to them these words penned by Warren Buffet in his letter to shareholders in 2013: “Forming macro opinions or listening to the macro or market predictions of others is a waste of time. Indeed, it is dangerous because it may blur your vision of the facts that are truly important.”