The Best Economic & Market Indicator In The World?
March 10, 2006

Is there an ultimate economic indicator that can be used consistently over time to predict stock market gains? Well, I'll always contend it's a pipe-dream, but we have found several pieces of economic data that are actually helpful for investors. That is, they're actually useful in helping make good decisions about when and how to invest (while ironically, most economic data is NOT helpful for investors). Some of this data is obscure, but some of is common and very straight-forward.

One of our favorites of the straight-forward variety is unemployment. Yes, that right - plain old unemployment. It's a bit unusual for us (of all firms) to use something so simple. However, our first goal is to make money for our clients. If using the obvious tools can do that, then that's what we'll do.

So what's the strategy?

It's bizarrely simple, yet very effective. When unemployment is sinking, stocks have historically been shown to rise at an accelerated pace. When unemployment is on the rise, stocks are usually weak, if not bearish. That's not earth shattering. It's not even a surprise. And more than that, it's the complete opposite of our ongoing argument that the economy is NOT the market. Yet it's one of the best tools that any investor can really use.

Our clients and regular readers will undoubtedly be wondering why the sudden conventional interpretation of economic data. After all, we have always been skeptical to do so (and have been well rewarded for not doing the obvious). The answer, in short, is that unemployment is not actually economic data. OK, to a tiny degree one could justify it as economic data....it has economic components. But by and large, it's corporate data. That's a key factor, since stock prices are usually set in relation to corporate profitability, and NOT economic factors.

Since 1970, there have been four major periods of rising unemployment. Each one was accompanied by a fairly unimpressive return for the S&P 500. They are as follows:

• Oct. 73 to May 75: Unemployment rose from 4.6% to 9.0%, while the S&P 500 falls from 108.28 to 91.14 (and had been as low as 60.96). That doubling of the unemployment rate meant at least a 15.8% dip for stocks.
• May 79 to December 82: Unemployment went from 5.6% to 10.8%, as the S&P 500 traveled from 99.08 to 140.63. However, the S&P 500 had been as low as 101.44 in September of 82, so the majority of that period was problematic for the period despite the 41.9% gain in the last three months of that term.
• March 89 to June 92: Unemployment rose from 5.0% to 7.8%, while the S&P 500 climbed from 294.87 to 408.13. That 38.4% gain was actually pretty impressive.
• April 2000 to June 2003: Unemployment went from 3.8% to 6.3%. The S&P 500 fell from 1452.45 to 974.50, and has been as low as 768.65. The net loss for stocks during that time was 32.9%.

The average market return, as measured by the S&P 500, in those four periods of rising unemployment was 7.9%. Two losses, two gains. An optimist might argue that any gain is still a gain, but in this case, the average 7.9% rise doesn't justify the 50/50 odds of getting it. Plus, given that the average rising unemployment rate term lasted just under 35 months, it's a very uncompelling argument. That makes the average annual gain in a rising unemployment environment a tepid 2.7%.

In the interest of fairness, let's also take a look at how the market responds when unemployment is pointed lower. These five periods are just the points in time that don't include the periods discussed above, again going back to 1970.

• August 71 to October 73: Unemployment moves from 6.1% to 4.6%, while the S&P 500 went from 99.02 to 108.28, and had been as high as 121.74. It was still a net gain of 9.3%.
• May 75 to May 79: Unemployment moves from 9.0% to 5.6%. The S&P 500 saw a mild gain of 8.7%, from 91.14 to 99.08.
• December 82 to March of 89: Unemployment slides from a modern-era high of 10.8% to only 5.0%, while the S&P 500 rallies from 140.63 to 294.87. That's a gain of 109.6%.
• June 92 to April 2000: Unemployment falls from 7.8% to 3.8%. The S&P 500 rose from 408.13 to 1452.45, for a heroic 255.8% gain.
• June 2003 to February 2006 (and still counting): Unemployment sinks from 6.3% to 4.8%, while the S&P 500 heads from 974.50 to 1272.25. That's a 30.5% gain, so far.

The average gain during periods of falling unemployment? A whopping 83.1%, with the current rally still bolstering that number. And even if you take out the monster rally from the late 90's, the average return was still 39.9%. That said, it would actually be bad practice to remove from this analysis the red-hot returns form the late 90's, as it indeed could happen again. These periods of lowering unemployment lasted an average of 50.4 months, and again, that number is being pushed higher by the current downtrend in unemployment. As you can tell, the longer unemployment sinks, the better for stocks.

The average annual return during a period of falling unemployment was 19.8%.

Compared to the average annual return of 2.7% when unemployment is on the rise, it can't just be a coincidence that things are so good for stocks when jobs are plentiful, and so bad when jobs are scarce.

This is one of our primary indicators we use for our managed accounts, and it helps enormously. That said, there are a couple of ongoing challenges that it presents.

First, it's not always easy to tell when the direction unemployment is headed is actually changing. Sometimes we see a slight rise when it's falling in the bigger picture (e.g. the rise to 4.8% in February 2006), or a slight drop when it's rising. Both can lead to fakeouts if you react too quickly. That's why we tend to wait at least two to three months before making that kind of call. We also factor in other data, to act as confirmation of any shifts in unemployment figures.

Second, it takes patience. This is monthly data, and it technically takes at least two months worth of data to spot any trend (and actually much more to spot a good one). It's very tough to sit tight and do nothing when it seems like everyone else is telling you have to do something right now. The fact of the matter is that, as we have seen, high-quality opportunities last much longer than a month. When unemployment was sinking, it did so for an average of more than 50 months. That's plenty of time to make some money. Waiting for a month or two before acting won't put you behind the eight-ball. It's just being prudent. Yet, it's still a mental and emotional challenge, even for us.

Day in and day out, we look at a lot of data in our effort to reap big returns. If we had to narrow that down to just one piece of data though, unemployment would be at the top of a short list.

And on a side note, the current unemployment rate of 4.8% is under both the long-term (1970) average of 6.2%, and the modern era average (1990) of 5.5%. I keep getting the sense that investors are worried that unemployment is 'too high', but it's really not. It's actually quite low, and clearly going lower. Quite frankly, my concern is that it's approaching levels that have been historically too low to be sustained. With the exception of the late 90's when all the 'norms' were redefined, we're seeing unemployment at levels we haven't seen since the early 70's. I can't help but wonder if the unemployment rate is actually ready to creep back up towards more familiar levels, which is not exactly the best scenario for stocks. Fortunately, interest rates are also still under long-term norms, which may be enough to keep investors in a bullish mode. But that's another story.

S&P 500 Versus Unemployment - Monthly