The Wall Street Journal featured a helpful article “Happy New Year…Now What?,” (subscription required) that stated:
Collectively, Wall Street forecasts are similarly inconsistent. Since 2000, the S&P 500 has returned a 4% average annual gain, excluding dividends. By comparison, Wall Street strategists have predicted 10% yearly returns, according to Birinyi. And, on average, the consensus always has predicted annual gains, missing all five down years in that stretch.
So what’s my take? My take is that Wall Street strategist forecasts are essentially useless.
Going into 2015, I wrote that SPX options had a skew to the downside, implying that there was more downside risk. Using a similar process, let’s look at SPX options for 2016.
On Thursday, December 31, 2015, the SPX closed at 2043.94. Looking at the SPX December 2016 Options (those that expire on December 16, 2016) and using thinkorswim’s platform for options, I note the following:
- About an 80 percent probability that SPX closes above 1725;
- About an 80 percent probability that SPX closes below 2275; and thus
- About a 60 percent probability that SPX closes between 1725 and 2275 at expiration.
I arrive at these values by finding the SPX value at expiration where the put delta is roughly -0.2 and the call delta is roughly 0.2. I am using option deltas as a rough proxy for the probabilities.
The lower level is about 84 percent of the current SPX value and higher level is about 111 percent. Again, we see a skew to the downside because that implies about an 18 percent downside risk compared to about a nine percent upside risk. There is an approximately equal risk that SPX closes below 1725 as there is that it closes above 2275.
Even if the risks were perfectly balanced, there should be a skew to the downside because of dividends. Dividends make puts more expensive. So let’s adjust the put value to incorporate the dividends. SPY dividends have been about 2.4 percent. 2043.94 times 2.4 percent yields 49.05. Thus, to arrive at the amount of skew while incorporating dividends, we get a higher level of 86.8 percent (equals (1725 plus 49) divided by 2044). Thus, the downside risk of about 15 percent is nearly twice as great as the upside risk of about 11 percent.
When learning of various forecasts, remember to keep an open mind. As the opening quotation shows, Wall Street forecasters are traditionally an optimistic bunch. While option pricing models are certainly not crystal balls, they do provide a quick snapshot as to how market participants view the current market. With the passage of time, of course, that view will change. In short, I am not wildly optimistic in my outlook for 2016.
Hey Kevin! Nice article. Have you gone back and done similar math on years past? It’s fascinating that strategist don’t seem to employ your method.
Regards,
Kim G
Hey Kim, Thank you for your comment. You asked if I backtested this methodology. Last year I used the same methodology but I haven’t gone back beyond that. One should remember, too, that there is still a 20 percent probability that the market ends lower than the lower endpoint. And, there’s roughly a 40 percent probability that the market touches or exceeds the lower endpoint at some point throughout the year. The same applies to the upper endpoint.
With the rough start to 2016, I suspect that if we were to do the same exercise again today, our endpoints would be even further apart. The market is lower now than it was at year end. So that alone suggests a possible lower endpoint. And, volatility is higher, which will widen the distance between the upper and lower endpoints.
You further mention that it’s fascinating that strategists don’t seem to employ this method. If they did, they wouldn’t have much of a job. It takes all of five minutes and anyone can do it. Furthermore, I suspect many would complain the range is so large that it doesn’t help them assess the market for the following year. I would counter that argument, however, by stating that it is better to have a large range than to consistently provide erroneous forecasts.
The US Energy Administration Information (EIA) does, however, use this methodology in its pricing forecast. When I look at the graph, I notice the “95% NYMEX futures lower confidence interval.” I see the upper confidence interval as well. We are employing the same approach but on different underlying assets.
We should all keep in mind that this is merely the probability range of what the market is pricing; anything could still happen. I rather doubt that 2008’s ex-ante range encompassed what actually did happen.
Cheers!
Agreed. This is just a probabilistic assessment. There’s no certainty and the probabilistic assessment will change throughout the year.