The current stage of the business cycle and the timing of the next recession.
The use of leading indicators to time the recession.
Timing the recession is not the same as timing the market.
How to position the portfolio for the current stage of the cycle?
Current stage of the business cycle
We are in the seventh inning of a current business cycle per Ray Dalio’s comment in September this year; putting the timing of recession to 2020. Slightly earlier yet (June 2018), Jamie Dimon noted that we are in the sixth inning, which would recession to mid-2021. If we look at yield curve (widely discussed topic among Seeking Alpha community members), 2s/10s spread is quickly approaching negative zone and might cross into the red within the next 6 months. You might recall that there is a correlation between yield curve inversion and recession, which seems to imply that first precedes the other by around 1-2 years. This all makes 2020 or 2021 likely candidates for the next recession.
According to Credit Suisse, the 2s10s curve has inverted prior to each recession for the past forty years. At the current level of 2s10s, the last forty years of data would suggest there’s less than a 50% likelihood that we’ll enter a recession within two years. Note that recently, 2s5s and 3s5s inverted; 2s5s inversion suggests ~60% likelihood of a recession within two years based on the last 40 years of data.
Figure 1. 2s10s (source: Credit Suisse)
Figure 2. 2s5s (source: Credit Suisse)
Federal Reserve Bank of Cleveland’s recession probability model indicates below 20% change. Note, however, FRB Cleveland model has hovered around 40-60% during recessions.
Figure 3. The probability of recession (source: FRB Cleveland)
There are other indications of the maturing cycle, such as tight monetary policy (Fed rate hikes, shrinking Fed balance sheet), and a tight labor market and increasing wage pressure.
Figure 4. The US and Global business cycles are maturing (source: Fidelity)
Just a reminder: correlation does not mean causation (i.e., the inverted curve does not cause a recession), and it is unlikely that anyone (even the big names in the investing world) would have a crystal ball. That said, consensus indicates that we are in the later stage of the business cycle.
Use of leading indicators to time the recession
Even though yield curve is considered one of the indicators of upcoming recession both by academic and practitioners, there is a long list of leading indicators for the U.S. published by the Conference Board (Leading Economic Index, later “LEI”):
- Average weekly hours (manufacturing)
- Average weekly initial claims for unemployment insurance
- Manufacturers’ new orders consumer goods and materials
- ISM® Index of New Orders
- Manufacturers' new orders, nondefense capital goods excluding aircraft orders
- Building permits (new private housing units)
- Stock prices (500 common stocks)
- Leading Credit Index
- Interest rate spread (10-year Treasury bonds less federal funds)
- Average consumer expectations for business conditions
LEI performance as a predictor of the upcoming recession (determined by NBER) was impressive:
- 1990 recession: LEI peaked in Nov 1988, which was 21 month prior to the start of the recession
- 2001 recession: LEI peaked in Oct 2000, which was 6 months prior to the start of the recession
- 2008 recession: LEI peaked in Apr 2006, which was 21 months prior to the start of the recession
Note that I’m qualifying such performance as “impressive” given that even NBER takes months before it determines the start and end of recessions.
Those who are interested in market timing would not only want to know when the economy turns sour but also when economic clouds lift. From that perspective, LEI has done decent job pointing out the end of recession:
- 1991 recovery: LEI turned positive in Mar 1991, just when the recession ended
- 2001 recovery: LEI turned positive in Oct 2001, which was 1 month prior to the end of the recession
- 2009 recovery: LEI turned positive in Mar 2009, which was 4 months prior to the end of the recession
But we will see in a minute that such impressive performance does not help market timers much.
Timing the recession is not the same as timing the market
LEI seems to be the ideal tool to use by a market timer, right? No, it is not! Timing the recession is not the same as timing the market. You might have noticed by now that performance of S&P 500 is one of the leading indicators. This, by definition, means that S&P 500 peak is expected to ‘lead’ / precede the start of the recession. So, let’s look at how good is LEI in predicting S&P 500 peaks and bottoms:
- 1990 recession: LEI peaked in Nov 1998, which was 20 months prior to S&P 500 peak
- 2001 recession: LEI peaked in Oct 2000, just when S&P 500 peaked
- 2008 recession: LEI peaked in Apr 2006, which was 20 months prior to S&P 500 peak
Hmm, someone who has observed that LEI peaks tend to align with S&P 500 peaks (based on 1990 and 2001 recession experiences) would have exited market on April 2006, which would have been 20 months earlier than S&P’s peak in November 2008 (Note that I’m using the S&P 500 total return index, which reflects the impact of dividends rather than S&P 500 price return which does not). During these 20 months, S&P 500 has provided a total return of 25%. Not ideal!
Now, let’s look at how LEI would “help” re-enter the market:
- 1991 recovery: LEI turned positive in Mar 1991, i.e., 4 months after S&P 500 started recovering
- 2001 recovery: LEI turned positive in Oct 2001, which was 12-month prior S&P recovery
- 2009 recovery: LEI turned positive in Mar 2009, just when S&P 500 started recovering
Similarly, if you would have relied on 2001 experience, you would likely 12 months after LEI started recovering to get back to stocks; during this period, you would have missed 54% gain in S&P 500. Overall, this is far from being impressive.
Also, don’t forget that it is impossible to know ex-ante whether LEI peaked looking only at LEI’s price chart. Additionally, don’t forget that there’s a lag between the time LEI is published and you can rebalance your portfolio. In practice, one would need to define some trading rules to backtest usefulness of LEI for a market-timer.
Here are the results of a few trading rules for the last 30 years (since January 1989; assumes monthly rebalancing, except for “buy-and-hold”):
$10K original investment
Buy and hold S&P 500
| || |
| || |
“LEI + price signal”
| || |
Table 1. Backtests of various investment approaches
As you can see from the table, following simple monthly LEI change as an indicator to go long or short stocks adds marginal excess return over ‘buy-and-hold’ approach: 10.9% vs. 10.3%. Note that the hit rate for LEI only (and other timing approaches) is relatively low. Instead of tracking LEI, one could have used a simple 12-month price momentum to get a similar result (11.1%).
While ‘buy-and-hold’ investors see 2/3 of the time (on monthly basis) increase in their equity line, market timers using the above approaches would see 2/3 of the time decrease in their equity line. How then such market timing approaches still provide an improvement over ‘buy-and-hold’? That’s because market timers avoid large cumulative losses during recessions that buy-and-holders absorb.
Now, let’s combine LEI and S&P price signal (“LEI + price signal” approach in the table), which yields 14.2% annualized return. This would mean over 2x absolute dollar return (thanks to the power of compounding) during these 30 years compared to other approaches in the table. Not bad!
Would I recommend using “LEI + price signal” approach”? No, I would not for the following two reasons:
- I am not providing any investment advice in this article. This article is only for educational purposes. Please consult your judgment and/or investment advisor before making any investment decision.
- Most investors would find 32% hit rate, i.e. 2/3 of the time having negative return months, psychologically painful/unacceptable. This would make this approach unsuitable for most of these investors.
Figure 5. LEI (LHS) an S&P 500 (NYSEARCA:RHS) (Source: Fed St. Louis)
How to position the portfolio for the current stage of the cycle?
Moving out of stocks into cash does not seem to be warranted at this point. In his Seeking Alpha article, Dilantha De Silvanotes that after 2s10s inversion S&P 500 (SPY) delivered 21% on average prior to the peak. And yet 2s10s haven’t inverted so far.
Below are few ideas about portfolio positioning for the late cycle (none of the below is investment advice; you might find a few more good ideas from Nassim Taleb’s and books):
- Nassim Taleb approach: allocate the majority of the portfolio (let’s say 80%) in short-term fixed-income instruments and rest (e.g. 20%) in calls (preferably LEAPs). This approach allows you to participate in the melt-up phase while ensuring that when the market moves into melt-down – your losses are limited to the size of your call options allocation.
- Colm O’Shea approach: go long FX straddles on cheap vol FX pairs (some asset class vols tend to get too cheap as market switches into euphoria) and short credit (credit spreads tend to widen as the stock market approaches ultimate peak; seems that fixed income markets tend to act as a leading indicator)
- Michael Platt approach: use systematic trend following, which is likely going to suggest reversing long equity and commodity positions
- Tom Claugus approach: short companies with BV>5x and CFO<0. Highly valued cash-burning companies tend to do particularly poorly during late stages as investors become pickier. You might want to also add one more attribute to your “short” screener: highly levered companies.
- Tactical asset allocation approach: go long EM (EEM) and commodities (DBC) as they tend to outperform during late expansion. Once curve inverts, get out of EM. Typically, commodities continue performing well after S&P peak (last time it was 9 months). Commodities tend to be good inflation and geopolitical hedge (e.g. oil and tensions in the Middle East).
- Buy-and-hold approach: this would be a viable approach for many retail investors as long as they can stick to the concept of long-term investing and are willing to see half of their portfolio wiped out during the market downturn.
Goldman Sachs notes that commodities are a “hedge against rising interest rates and are one of the few assets which outperform in the late-cycle”. While financial markets depend upon growth rates, commodities depend upon activity levels. In the late-cycle, growth rate decelerates driven by increasing interest rates. Activity levels, on another hand, remain high even despite slowing growth.
Once again, the list above is just a few thoughts for your consideration. Please share your views in the comments below and any other investment approaches/ideas for the late cycle.
Disclosure: I am/we are long EEM, DBC. 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.