By Jerry Wagner
Weather forecasts have forever been maligned for their inaccuracy - hence this week's title, a quote often attributed to Mark Twain. But it seems that they have become increasingly accurate now that we have satellite imaging and have given up on long-term forecasting.
You can now check the accuracy of your favorite weather provider online. I did recently and found the rankings were pretty darn good. The accuracy of the next-day and 10-day forecasts of rain were somewhere between 75% and 85%.
Of course, these are pretty narrow time windows. As journalist Miles Kington stated some time ago, "Satellite photography in the 1970s gave rise to the long-range weather forecast, a month at a time. This, in turn, gave rise to the observation that the long-range weather forecast was wrong most of the time. In turn, this gave rise to the dropping of the long-range weather forecast and to the admission that really accurate forecasting could only cover the next day or two, and not always then."
One of those "not always then" moments in forecasting history occurred last weekend. I had expected some "outside time" on Sunday. The forecast was for only a 24% chance of rain.
You guessed it - it rained. Checking the radar map, the rain cell was just a blip on an otherwise clear screen for my entire state. But for me, it did not matter. I was stuck in a deluge.
And I can't really blame the weather forecasters. They gave me the odds, and, as in most things, the chance of an occurrence was not 0% or 100%. It was someplace in between.
I'm not sure whether most of us truly understand statistics. They get used routinely, but do we really understand the provisos that go with their use?
We speak of the average temperature of a place but don't seem to realize that the high and the low that are a part of that statistic can be very far apart. The average temperature in Denver is 50 degrees, but it ranges from a monthly average of 17 degrees to 88 degrees. Of course, on a daily basis, the range is much greater. Similarly, the average return in the stock market has been about 11% per year since 1928, but that includes a gain of 52.56% in 1954 and a decline of 43.84% in 1931.
Probability gives us all kind of problems. The probability of the stock market rising on any single day is about 54%, but on days that it doesn't go up, the one-day loss has been more than 25%. Remember October 19, 1987?
You have to take statistics - and especially forecasting - with a grain of salt. They can tell us what is likely, but rarely what is certain. This is almost always the case unless your forecast is of the George Carlin variety: "Weather forecast for tonight: dark!" (Of course, even that can be wrong if you're one of those residents living near the Arctic Circle!)
As with the weather, financial forecasting tends to be more accurate for short-term rather than longer-term time periods. Forecasting out a year in advance can be just plain crazy, but Wall Street does it all of the time.
Generally, the best advice is what Nobel laureate Niels Bohr voiced, "Prediction is very difficult, especially if it's about the future."
Another quarter is now in the books, and it differs greatly from the first quarter of the year. At the end of the first quarter of this year, both the Dow 30 and the S&P 500 Indexes were down. At the end of quarter two, all of the popular U.S. equity indexes posted solid gains, with the S&P up better than 3%.
Gold also reversed direction, but that was not a good thing. While it had been up for the first quarter, it fell just over 4% in the second quarter. Bonds were consistent; they generally fell in both quarters!
The increase in volatility that we observed in the first quarter versus the low volatility of 2017 continued. The daily percent change of the S&P 500 changed by more than 1% only 3.2% of the time in 2017, but more than 29% of the time so far in 2018. Similarly, the absolute daily change in the Index has more than doubled.
And it has been quite a daunting year on the interest-rate front. Leading the Federal Reserve ever higher, the bond market has taken interest rates in a pretty steady upward direction. This is in contrast to the general decline that we experienced last year that supported both higher stock and bond prices.
Higher interest rates and volatility are not a prescription for higher stock prices. The tendency is the contrary.
Stock market valuation has become less extreme as a function of the increase in earnings. Earnings were 24% higher in the first quarter of the year and are expected to average 20% higher for the rest of this year.
Pre-midterm elections can mean a flat summer in terms of stock market returns when the market is up modestly for the year, as is the case this year. At the same time, July has generally been an exception to the summer blahs in the market, gaining ground on average over the last 100, 20, and 10 years. The July gains, however, have usually come after the first week to 10 days.
One final thought on forecasting: The great 20th-century economist, reputedly a phenomenal investor, John Kenneth Galbraith, is said to have remarked, "The only function of economic forecasting is to make astrology look respectable." I guess that pretty well sums it up.
Enjoy your holiday-shortened week.