Never saw it coming - MoneySense

Never saw it coming

Forecasters have an ugly little secret – they can’t actually predict anything.


The forecast season is upon us. This year, as every year, economists will line up in December to predict what will happen in the 12 months ahead. These superbly educated experts will consider all the data and draw on years of experience to tell us where the economy, employment, oil prices and the stock market are headed next.

Oh, but there is one hitch — you shouldn’t count on those forecasts to be all that accurate. In particular, don’t count on finding out what matters most to Canadians this year — when the economy will turn around.

The unfortunate truth about economic forecasting is that it is fundamentally unreliable. So if you’ve been reading the experts’ predictions, and thinking about making major changes to your portfolio as a result of what you’ve read, think again. The failure of most forecasters to foresee the recent downturn is nothing unusual. “Economists are simply hopeless when it comes to forecasting recessions,” writes James Montier, an economist at Société Générale in London who has written extensively about forecasting. “Actually, I could have stopped that sentence before the word recessions.”

Smart investors don’t forecast; they anti-forecast. They realize how uncertain the future is and put little faith in anyone’s predictions. Rather than using forecasts to shape their portfolios, they use them to ensure that their portfolios aren’t blindsided by unseen risks. Here’s how to make anti-forecasting work for you.

Anti-forecasting begins with the proposition that nobody can predict the future. At first this seems hard to believe. Why would banks and brokerages employ so many staff to produce forecasts that don’t pan out? It’s an excellent question — and the answer may be that if you’re running a bank or brokerage you have to act as if you can predict the future, because clients expect it. Fortunately for financial institutions, most clients don’t keep tabs on how forecasts pan out. If they did, clients might be dismayed.

Flash back to December of 2007. Most economists in Canada and the U.S. predicted the economy would stride ahead in 2008, growing by 2% to 3%. They saw equally good or better prospects in store for 2009. The forecasters believed home prices in Canada would either stabilize or make small gains. While sliding home prices in the U.S. were recognized as grounds for concern, the forecasters thought other factors, such as growing U.S. exports because of a cheaper greenback, would provide enough boost to offset the real estate downturn. Typical of the measured optimism of the time were comments by Jeff Rubin, chief economist and chief strategist at CIBC World Markets: “A two-quarter slowdown, contained by more Fed rate cuts, should set the stage for a gradual re-acceleration of economic growth over the course of 2008, and a subsequent rally in North American stocks to new cyclical highs.”

Look what actually happened. The Canadian and U.S. economies slowed in early 2008. U.S. investment banks crumbled. The downturn in U.S. home prices gained speed and turned into outright collapse. Global stock markets crashed. By late fall, it was clear that the U.S. was in recession and the world was in the grip of the biggest financial crisis since the Great Depression. While it had not been confirmed that Canada had fallen into recession as we went to press, home prices had begun to fall, the Toronto Stock Exchange was in freefall, and Stephen Harper was vowing to fight the downturn.

The most candid forecast during this dreadful period came in October from the University of Toronto Institute for Policy Analysis. As markets crashed, it issued a report with the headline “We Don’t Have a Clue and We’re Not Going to Pretend that We Do.” Forecasters Peter Dungan and Steve Murphy wrote in the report: “The short-term outlook at this point depends on a huge number of either/ors, some of which we do not even pretend to understand, never mind try to forecast.”

Forecasters are rarely so honest about their predicting abilities. Rather than confess to ignorance, most will continue to make forecasts, adjusting them as necessary. After all, that’s what economists are paid to do, despite a marked lack of success.

Forecasters have a particularly bad record when it comes to calling turning points in the economy. Most of them failed to predict the extent of the dotcom crash at the turn of the century. Most failed to identify the dire repercussions of the U.S. housing bubble. And most also failed to predict any of the last three U.S. recessions, based on evidence from the quarterly survey of economic forecasters by the Federal Reserve Bank of Philadelphia. “I don’t think we, as a profession, ever had an ability to forecast recessions,” Jeffrey A. Frankel, professor of economics at Harvard University told The New York Times in 2007. “It’s hard enough to know when a recession has started, looking at it with hindsight.”

Even when the economy is growing, forecasts are blurry. Steve Murphy of the University of Toronto Institute for Policy Analysis and Bryan Campbell of Concordia University reviewed Canadian economic forecasts from nine banks and other organizations over the period from 1984 to 2003. They found that the average of the annual forecasts for growth in gross domestic product missed the actual result by an average of 1.3 percentage points. That is, if the economy grew by, say, 3%, the average forecast would typically have been 1.7% or 4.3% — not much help at all if you’re trying to plan your investing strategy based upon those forecasts. Why is the economy so hard to predict? At any one time, there are multiple forces pulling it in different directions and it’s hard to know which forces will prevail. To make the challenge even more difficult, collecting numbers takes time and forecasters are often working with data that is months old.

But the problem extends beyond data or economic models. The future seems to have an endless capacity to surprise us, and it’s not just economists who get their forecasts wrong. Political observers and social commentators also mess up. One of the most comprehensive studies of forecasting accuracy was performed by Philip Tetlock, a professor of business and political science at the University of California at Berkeley. Beginning more than 20 years ago, he selected 284 people who make their living offering advice on political trends and asked them to assess the probability that various events would occur: would Canada split up?; would the U.S. go to war in the Persian Gulf?; would apartheid end peacefully?; and so on. All told, the experts made more than 82,000 forecasts by the time the study ended in 2003.

When Tetlock analyzed the results, he came to a surprising conclusion. Experts can’t predict broad trends any better than a random guess would. And specialists in an area can’t forecast more accurately than non-specialists. Yes, subjects who know a little about an area can forecast better than someone who knows nothing, but knowing a lot doesn’t seem to improve forecasting accuracy. Forecasters seem to rapidly reach a point where adding more knowledge doesn’t make predictions any more reliable.

The diminishing returns from knowledge have been particularly evident in the financial crisis of recent months. While the stock market crash took small investors by surprise, it was just as big a shock to most economists — including the ones at Lehman Brothers. They thought things were fine, right up until they weren’t. As the giant investment bank was declaring bankruptcy in September, its staff economists sent out a poignant commentary in their last Lehman Brothers weekly economic letter: “The episode of financial crisis appears to be much deeper and more serious than we and most observers thought it likely to be.”

To their great credit, a few lonely economists did foresee the current crisis. Among them is Robert Shiller, professor of economics at Yale, who has long argued that the U.S. housing boom was a bubble. Nouriel Roubini, professor of economics at New York University, also got the big picture right. He predicted this past February that one or two major U.S. financial institutions might go belly-up. Most people thought he was an alarmist, but his only miscalculation turned out to be underestimating the number of failures. As it turned out, at least nine major U.S. financial institutions were forced out of business in 2008, counting forced takeovers at fire-sale prices.

So can you prosper in the years to come by listening to one of these far-sighted gentlemen? Maybe — but it all depends on how patient you are. Both Shiller and Roubini have argued since at least 2005 that a crisis was in the making. In 2006, Roubini predicted a recession — in 2007. If you had reacted to their dire warnings in early 2005 by taking all your money out of the stock market, you would have been sitting in cash for almost three years and missed double-digit gains in the stock market before it peaked in late 2007.

Shiller and Roubini prove that even the best forecasters usually get things right only in the long term. When it comes to predicting what the next few months will bring, only the foolish or the brave need apply. In fact, economists like to joke that if you really want to be right about your forecast, you should restrict yourself to predicting what will happen, or when something will happen, but never both.

The danger of being too precise in your forecasting is that you’ll wind up like Jeff Rubin, the CIBC economist. Rubin, a noted bull on oil prices, raised eyebrows in October 2007 by predicting that oil would bolt from $70 a barrel to $100 a barrel by the end of 2008. He also predicted that Canada’s commodity-heavy S&P/TSX composite index would hit 16,200 by the same date. Both forecasts were considered bold at the time.

When oil prices surged to over $140 a barrel by June, Rubin looked like the smartest guy in the room. He responded by raising his forecast for average 2008 oil prices to $125 a barrel. He went even further out on a limb and predicted that oil would hit $150 a barrel in 2009 and $200 a barrel by 2010. But then the roller-coaster turned downward. Oil prices began falling and Rubin began reforecasting. By late October, oil had fallen to below $70. Rubin was no longer talking about $200 oil in 2010 but rather within four or five years. Meanwhile he had progressively cut his end-of-2008 stock market forecast from his original call of 16,200. His prediction became 14,500, then 13,000, then 9,500 — at which point, most people had long since stopped paying attention to his forecast.

Oil prices are volatile, and Rubin remains a long-term bull on oil prices, so it’s possible he may have the last laugh. But for now he serves as a classic case of how rapidly forecasters can go from being absolutely right to being way off base.

So what good are forecasts anyway? If you see them simply as advice about where to place your investments for the year ahead, not much. But smart investors can still make good use of forecasts. Here are some of the best uses:

To think about the big picture: Forecasts can help you think through what is likely to happen over the long term. Investors who read Shiller and Roubini’s work in 2005 and 2006 could have taken into account the possibility that the housing market would crash. Those investors may not have wanted to flee to the safety of cash, but they might have hedged their portfolios against the possibility that Shiller and Roubini were right by adjusting their holdings away from the sectors — such as banks and construction companies — that were most exposed to a real estate downturn.

To detect unseen risks: Whether you’re a bull or a bear, it helps to have your assumptions challenged. It pays to read widely, especially among forecasters who have outlooks different than your own. It also helps to read forecasters who come from a variety of backgrounds — universities or research centres, for example, as well as banks and brokerages. You may ultimately decide to ignore these conflicting opinions, but they may also open your eyes to factors you haven’t considered. For instance, someone who read Rubin’s optimistic forecasts about oil prices a couple of years ago would have learned a great deal about the galloping increase in global demand for oil and the declining state of most production fields. While his arguments may not have convinced you to invest in the oil sector,they would have helped you to understand that the future for oil prices was likely to be a lot more volatile than most people figured.

To spot growing pessimism: Economists realize that the economy grows in the vast majority of years, so they’re reluctant to go against the odds and predict the relatively rare event of a recession. But they are quite willing to acknowledge the heightened probability of one. Thus, while the consensus of U.S. economists failed to predict the last three recessions, U.S. recessions often occur after economists think the chances of the economy shrinking in the next quarter are at least 30%. Knowing this, you can put yourself on alert if you see a growing number of economists raise their probability of a recession past the 30% level.

To realize the limits of knowledge: Perhaps the greatest benefit to reading a lot of forecasts is that you begin to realize that nobody — not economists, not stockbrokers, not investment pundits — really knows how things will turn out. As a result, you can begin to plan for many possibilities, not just one particular version of how the future may develop. Rather than locking yourself into one version of what may happen, you can prepare yourself for whatever may come.

Think of this as all-weather financial planning. It’s like having all-weather radials on your car. While all-weather radials won’t perform quite as well as slick tires in good weather or as snow tires in snowstorms, they perform reasonably well in all conditions. In much the same way, an all-weather portfolio is never going to be the top performer in any given market situation, but will perform decently no matter which way the stock market jumps.

Your all-weather financial plan should include conservative fixed-income investments such as investment-grade bonds or government-insured GICs. These products don’t earn a lot of interest these days, but they provide rock solid protection if the economy weakens. Your all-weather portfolio should also include stocks, because over the long haul, stocks are likely to recover from their current malaise and provide you with superior returns. Stocks can also help protect you from inflation, because their earnings and dividends tend to rise in tandem with rising prices. A good all-weather portfolio should span stocks from many different sectors and many different parts of the world. (For an easy-to-follow approach to building a diversified, low-cost, all-weather portfolio, visit and look up Couch Potato Investing.) Once you have your all-weather portfolio in place, you’re prepared for whatever the future may hold — which means that you can stop worrying about the experts’ latest outlook for the future and get on with the far more interesting business of enjoying the here and now.