Low Jobless Claims Likely Mean Future Market Losses

|
Includes: SPY
by: Michael Strohmann

Summary

Jobless claims are currently at some of the lowest levels in decades.

Investors should not assume that low jobless claims mean that markets will be strong going forward.

Large spikes in jobless claims frequently occur when claims are far below their average and these spikes are associated with steep market losses.

Last week, I wrote about Treasury spreads and their implication for future market gains or losses. This week, I want to look at jobless claims and whether or not they can tell us anything about the future market direction. Specifically, I want to investigate if jobless claims can warn us of large, future market losses.

This past week, the Labor Department released its weekly jobless claims and they came in lower than expected. This prompted some pundits to declare that a recession and associated market downturn were now less likely. If layoffs are lower than expected, then the economy seems relatively healthy. This line of reasoning makes sense, but fails to take the cyclicality of the economy into account.

Some History…

Let's first look at the historical jobless claims(4-week moving average):

Click to enlarge

Looking at the graph above, some things stand out: First, jobless claims, though having quite a lot of variance, do not show much of a trend over time. Second, jobless claims are mean reverting. Finally, jobless claims spike during recessions/market drops.

Anyone looking at the above graph can see that large increases in jobless claims happened quite suddenly and often happen when claims are below their historical average. That is the situation we find ourselves in today.

The Model

To quantify the likelihood of market drops, I used a series of logistic regressions. I took the percent that claims were above or below the long-term average and compared them against a binary variable: 1 if a future market loss >10% occurred within a certain time horizon and 0 otherwise. I then ran these regressions for 8 different time horizons. The results are shown in the table below:

Click to enlarge

The above table summarizes what we saw in the first graph. The lower the claims rate is below the long-term average, the higher the likelihood of large, subsequent market losses, as represented by the S&P 500 (NYSEARCA:SPY), in the following 24 - 48 months. Over any given time horizon and at any given claim rate, the chance of a drop greater than 10% in the S&P 500 is between 5 and 15%. Historically, when claims have been around -25%, as they are today, the chance of a major drop jumps to between 30 and 50% over an 18 to 48 month time horizon.

Side-note: Last week, I used the Wilshire 5000 as a proxy for the market and it is a better proxy for the market than the S&P 500. However, jobless claims data goes back to the 1960s and I do not have Wilshire 5000 data back that far - I do for the S&P 500. The truncated Wilshire 5000 dataset yields similar results to those above.

Quick Thoughts

  • There is no law that says jobless claims have to spike dramatically every 7-10 years or that, if they do, a major market drop will coincide with it. However…
  • …when predicting the future, history is as good of a guide as any and the historical data shows that jobless claims are highly cyclical and prone to jump dramatically from levels far below the average. Moreover, these jumps are often accompanied by steep market losses.
  • Though I'm bearish regarding the stock market, I'm open to new data convincing me that I'm wrong. Current jobless claims are not that data. I could accept that the labor market will remain relatively strong and we will not see any major market losses soon, however, nothing in the historical data tells me that a sustained market upswing is imminent.
  • It is probably time to get out of broad, market-based funds or ETFs. I believe holding cash for now is fine, but if you are uncomfortable with that, then it is time to start researching individual stocks that have solid dividends and strong balance sheets.

Conclusion

Though the historical relationship between below average jobless claims and future market drops may break down in the future, I do not see any reason to believe that investors can rest easy about the economy due to recent jobless claims. I believe markets will continue their downward trend due to flattening treasury spreads, rising credit yields, and declining earnings.

Data

The data used in this analysis was taken from the St. Louis Federal Reserves' FRED database. The data goes back to 1967 and can be found here. The code (Python) used for this analysis is also here.

Disclosure: I am/we are short SPY.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: I am very much hedged against SPY. While I have some long positions in these funds, I have even more long term put options.