The most financially destructive events investors face are recessions. By some estimates, more than 12 trillion dollars was lost during the 2008 Financial Crisis. In 2019, most people still quote a handful of arcane models that use a single piece of data to measure recession risks. With new innovations disrupting the status quo across all industries, why is there not a new and improved measure of recession risks? We have one.
Our goal is simple – to help investors make better decisions and protect their investments.
Often referenced recession models, like the one maintained by the Federal Reserve Bank of New York (The Yield Curve as a Leading Indicator) use the yield curve as a measure of recession risks. The yield curve is the difference between short-term and long-term rates. It also has a good track record of predicting recessions, however, there are limitation that even the NY Fed points out here (The Yield Curve as a Leading Indicator: Some Practical Issues).
Simply put, if you’re trying to avoid something as big and devastating as a recession, are you really going to pin your hopes on a single measure of the economy? It doesn’t make sense. Keep in mind, these models were originally developed back in the early 1990’s. It’s been nearly 40 years and people are still using the same models.
This is particularly important today as the yield curve recently inverted and many economists and pundits are talking up the possibility of a recession. What should investors do? Hunker down and ride it out? Sell before it’s too late? Before you decide to sell and put your money in the mattress, maybe we should consider a new approach to measuring recessions risks.
I worked at a large bank during the 2008 Financial Crisis. My employer was acquired by a competitor and the ensuing lay-offs lasted for years. They called it “operational efficiencies”, but that simply meant that half of us were redundant and unnecessary. I was one of the lucky ones, but nearly all of my colleagues were displaced in one way or another.
Following the crisis, I wondered if a better recession model could be built. If I was going to suffer through another recession, I want to see it coming well in advance. I spent thousands of hours collecting and analyzing economic data. Which pieces of data lead, which lagged, which combinations work once combined in different ways.
After years of research, I developed our Business Cycle Indicator. It is a new and improved recession model that doesn’t rely on just one piece of data. Instead, this model uses six different components. It can be modeled back to 1968 and would have predicted each of the 7 recessions that occurred since. Not only does it warn when a recession is likely, it also provides a signal when a recession is over.
Sound too good to be true? It’s just data. Here is a look at what goes into our model.
Before we get into the individual data points that go into our model, it’s worth mentioning the criteria we used in selecting the data. First, we only used data that spanned at least three business cycles. If there wasn’t enough historical data, we didn’t use it. Second, for those that had enough data, we analyzed its success rate at signaling a recession and the number of false positives. A model is useless if it warns of a recession and one doesn’t occur. In fact, it is detrimental.
Lastly, we looked at data that provides an opposing signal at the beginning and end of a recession. Meaning that the same piece of data needs to provide a signal before the beginning of a recession, but also a discernable signal when the recession is over. It is no good to have a model that tells you when to be more defensive, if it can’t tell you when to be less defensive.
The data that we selected spans measures of monetary policy, investment grade credit spreads, high yield credit spreads, labor market data, construction activity, and vehicle sales. Our Business Cycle Indicator uses a derivative of each data series and merges them together into a single data point. The indicator is updated monthly, with preliminary readings available at the end of each week.
The chart below looks at the history of our Business Cycle Indicator starting in 1968. Prior to each of the seven recessions that occurred since, our Business Cycle Indicator reached a level of -0.9 (red dotted line). Additionally, as the economy exited a recession, the indicator reached a level of +0.9 (green dotted line).
The table below shows the months in which our business cycle indicator warned of a recession, compared to when the recession officially started. It also shows when our indicator suggested the recession was over, and when the recession officially ended. On average, our indicator provided a warning eight months before the beginning of a recession, and six months after a recession ends.
Keep in mind, the actual announcement for the beginning and end of recessions happen with a significant lag. The National Bureau of Economic Research maintains records for these announcements (here). For example, the official announcement of the beginning of the Financial Crisis came in December of 2008. That was after Lehman went bankrupt and the S&P 500 was already down 42% from its all-time highs.
The official announcement that the 2008 recession had ended came in September of 2010. A full 18 months after the stock market bottomed and 14 months after the recession ended. Contrast that with our indicator’s expansion reading (+0.9) that came in July 2009 – right on time.
Now, let’s take a look at the history of the S&P 500 over the same time period – 1968 to present. The chart below shows the S&P 500 Price Index (excluding dividends). The periods that are shaded in blue are when our Business Cycle Indicator is in expansion. The areas shaded in red are periods when our indicator fell below -0.9 suggesting a recession, and had not yet reached +0.9 signaling a new expansion has begun.
Next, we take the above chart and breakdown the rates of return for the different periods – expansion periods, recession periods, and all periods. As one would expect, the vast majority of returns occurred while our Business Cycle Indicator was signaling expansion.
In fact, 187% of all the gains for the S&P came while our Business Cycle Indicator was in expansion (just 72% of all months). Those returns assume that you were simply sitting in cash earning nothing for those 175 months (14 ½ years) that our indicator was pointing to a recession.
It’s unrealistic to think that you would sit in cash for a cumulative of 14+ years during that time period. Let’s assume you bought the S&P back at the closing value on November of 1968 at 108.37. You held your position until our Business Cycle Indicator warned of a recession, at which time you sold and bought a 1-year treasury bill. You simply collected the coupon on your treasury bills until our indicator reached +0.9, at which point you go back to the S&P 500. You continue this process all the way to March 29th2019.
Obviously, had we simply held our one share of the S&P 500, our $108.37 would have appreciated to $2,834.40. However, our strategy that switched between the S&P and 1-year treasury based on our Business Cycle Indicator would have appreciated to $10,163.10. Not bad.
Being able to forecast a recession is important, but the hidden value in our Business Cycle Indicator is its ability to help investors stay invested. During this now 10-year-long bull market there have been some scary headline events that surely scared some investors out of the market. Here’s a short list off the top of my head:
- 2010 – Greek Debt Crisis – Overly indebted Greece to abandon the Euro, austerity.
- 2011 – European Debt Crisis – remember the acronym “PIGS” the troubled European Nations that were going to cause the Financial Crisis 2.0. They were Portugal, Italy, Greece, and Spain. The S&P 500 declined roughly 18%. Ultimately, this was a buying opportunity because there was no recession in the US.
- 2015 – Growth Scare – Oil prices plummeted crippling the energy sector leading to an earnings recession. No economic recession, so this was a buying opportunity.
- 2016 – Election – I hate to say it, but there were some people sold because of the outcome of the election.
- 2018 – Most recently in December of last year, we were on the cusp of entering a bear market. Markets had lost confidence in the Fed, we had Trade Issues, falling oil prices, you name it. Let’s face it, if you were reading the headlines, it was pretty ugly. However, it too was a buying opportunity.
- Today – The yield curve inverted last week, which should all but seal the fate of this economic expansion. Yet, stocks are rising.
In each of these instances, our Business Cycle Indicator never reached a level that suggested a recession was a risk.
Most people think the value of our Business Cycle Indicator is its ability to protect from events like the Financial Crisis or the Dotcom Bubble. That is true, but those events don’t come around very often. The hidden value in the model is the confidence it provides, knowing a recession is not around the corner, to remain invested or buy more.
Now, the logical question, where is our Business Cycle Indicator now. The chart below shows the last 12 months of our Business Cycle Indicator. While it did dip considerably in December, it bottomed at a reading of -0.4, quite far from the -0.9 that suggests a recession is eight months away. Today, the indicator has bounced back and is hovering near zero. This suggests that this business cycle still has legs (for now).
A more detailed explanation can be found on our Business Cycle Indicator page.