The flattening and inverting of the yield curve has been a major issue of discussion lately, as market participants dispute whether the latest inversions are a harbinger of a potential recession on the horizon. Flattening/inverting in the yield curve is perceived to have negative impacts on the broader economy, and consequently equity market performance as well. Nevertheless, statistical analysis finds that the S&P 500 can continue to perform well while the yield curve is flattening/ inverting, though its ability to foretell a recession should not be underestimated.

While the flattening/inverting yield curve has become more concerning lately as we approach inversions at critical levels, the curve has actually been flattening for years, during which the S&P 500 has continued to soar higher. In fact, statistical analysis on historical data finds that stocks actually tend to perform well during periods when the spread between the 2yr and 10yr Treasury yields narrow and eventually invert (where the 10yr yield falls below the 2yr yield).

I have analysed historical data for periods since the 1980s during which the 2yr/10yr spread narrowed and eventually inverted (from peak to trough), and have calculated the Pearson’s correlation coefficient ‘R’ between the spread and the S&P 500 during these periods. In the table below I have aggregated the results for each historical period. A value between 0 and +1 indicates a positive correlation, whereby the closer ‘R’ is to +1, the stronger the positive correlation. On the other hand, a value between 0 and -1 indicates a negative correlation, whereby the closer ‘R’ is to -1, the stronger the negative correlation.

Sources: the 2yr/10yr historical spread data was sourced from Federal Reserve Economic Data (FRED), and the S&P 500 historical price data was sourced from Yahoo Finance.

Evidently, there is a negative correlation between the 2yr/10yr spread and S&P 500 performance for each historical time period, with the strongest negative correlation occurring between 2003 and 2006. This implies that during periods when the 2yr/10yr spread has declined/ narrowed, the S&P 500 has historically continued to climb higher, even after the spread had fallen into negative territory (inversion occurred).

Now looking at the present time period, the spread between the 2yr and 10yr Treasury has actually been narrowing since December 31, 2013 from 2.66% to currently standing at about 19 basis points (at time of writing). While it has not inverted yet, certain investors fear that such an event would wreak havoc on the S&P 500. Nevertheless, the Pearson’s correlation coefficient ‘R’ between the 2yr/10yr spread and the S&P 500 for the time period between December 2013 and now is -0.87, once again implying a strong negative correlation. Moreover, historical data has revealed that the S&P 500 can continue climbing higher even as the yield curve inverts at this critical section.

To further assess the negative correlation between the 2yr/10yr spread and the S&P 500, I have conducted a linear regression analysis for the time period between December 2013 and now, to gain more statistical insight into the significance of the inverse relationship between these two variables for investors. Below you can find the scatter diagram and linear regression model for this data generated using StatPlus.

Note: This regression analysis only incorporates one predictor variable, the 2yr/10yr spread, given that I am only interested in the relationship of this specific variable with S&P 500 performance; though keep in mind that other variables also influence equity market performance.

Below I have summarized some key regression statistics results (using StatPlus):

The R- Squared value tells us how well the data set fits the linear regression model. For our specific data set (2yr/10yr spread vs. S&P 500), an R-Squared value of 0.75218 was obtained, which I would interpret as the data fitting the model relatively strongly. The Predicted R- Squared came in at 0.75154. While this is also relatively strong, I do not believe it is strong enough to conduct effective extrapolation (predicting future S&P 500 values using new observations for the 2yr/10yr spread), and that more independent variables are needed to effectively predict future performance. Nevertheless, my analysis is solely focused on assessing the relationship between the 2yr/10yr spread and the S&P 500, and while the model may not be strong enough to allow investors to predict future index performance, additional statistics for the model still strongly support the validity of the notion that there is a negative correlation between the two variables.

The F-test is another method to assess how well our data set fits the regression model through hypothesis testing. More specifically, it tests how well the data fits our model compared to an intercept-only model (a model with no predictor variables). Hence, the null hypothesis for our F-test is that the data fit for our model and the intercept-only model are equal, which we would ideally want to reject to find that our model fits the data much better than an intercept-only model. Having conducted this F-test with a significance level of 1%, I obtained a p-value of 0. A p-value result lower than the significance level, in this case 0.01, allows us to reject the null hypothesis. Therefore, this allows us to conclude that based on the results of the F-test with an extremely low significance level of 1%, our regression model holds notable significance in manifesting the inverse relation between the 2yr/10yr spread and the S&P 500, and thus further supports the validity of the notion of a negative correlation between the two variables. What does this mean for investors? Don’t place bearish bets on the S&P 500 based solely on a narrowing 2yr/10yr spread, as there is significant statistical evidence revealing that the index can continue climbing higher while the yield curve is flattening/inverting.

Does an inverted yield curve foreshadow doom for investors?

While we have found that equity market performance tends to remain favorable during periods when the yield curve is flattening and inverting at the 2yr/10yr section, keep in mind that such inversions tend to presage recessions. Hence, even though S&P 500 performance remains strong during the inversion, an economic downturn certainly does not bode well for stocks given the negative implications it has on corporate earnings/underlying fundamentals. Each of the historical 2yr/10yr yield curve inversions I covered earlier had been followed by a recession within a year of hitting their troughs. In the table below I have collated S&P 500 performance data during the previous three recessions.

Source: S&P 500 performance calculated using data from Yahoo Finance

It is interesting to find that stocks do not necessarily decline during every recession, given that the index yielded a positive return during the early 1990s recession. However, it is also worth noting that while returns were positive during that recession, a major crash in the stock market had already occurred in 1987, where the S&P 500 fell over 20% on October 19, and furthermore, the index had collapsed by 33.51% from peak to trough in the second half of 1987. Hence, the equity market had already taken a big hit during the years leading up to the recession. Nevertheless, investors should keep in mind that all three of those historical inversions in the yield curve have been followed by a recession.

Though various market experts claim that this time is different, and that an inversion of the yield curve at the 2yr/10yr section will not be followed by a recession this time. One of the main reasons experts claim an inversion this time would be a less reliable signal for a recession is that the ultra-loose monetary policy actions taken by the Fed post-crisis, including three rounds of Quantitative Easing (QE) and Operation Twist, have brought long-term yields artificially lower, which allows for flattening/inversions to take place more easily. However, I do not buy this argument. While the Fed did indeed take various measures to push long-term yields lower to support the economy following the Financial Crisis, they had also pushed short-term rates/yields to extremely low levels to a similar extent. The table below compares the 2yr/10yr spread peak prior to the previous three recessions to the latest peak in December 2013, and also exhibits what the 2yr and 10yr yield levels were on the dates that the spreads peaked.

The table above exhibits that the spread had peaked at levels between 2.60% and 2.75% before they started falling into inversion territories prior to the last two recessions. Similarly, the spread had reached 2.66% in December 2013 before it started falling down to current levels. So it was not as if the 2yr/10yr spread had become abnormally narrow amid the Fed’s ultra-loose monetary policy conditions, as it had peaked at a similar level relative to previous peaks prior to the last two recessions. Furthermore, the two columns on the right in the table above also reveal the parallel shifts lower in the 2yr and 10yr yields overtime since the previous spread peak dates. We find that the 2yr and 10yr yields fell by about equal magnitudes from their levels in July 2003 (which marked the previous spread peak) to December 2013. Hence, the argument that yield curve inversions are less reliable recession signals amid the belief that flattening/inversions take place more easily following the Fed’s ultra-loose monetary policy actions is invalid, given that the 2yr and 10yr yields have made parallel shifts lower of about equal magnitudes relative to each other. Therefore, from this perspective, investors should not underestimate the ability of a potential 2yr/10yr inversion to foreshadow a recession ahead.

Furthermore, ignoring the importance of the 10yr yield falling below the 2yr yield becomes even more irrational considering that the Fed has been reversing course on its ultra-loose monetary policy actions, by engaging in the Quantitative Tightening (QT) since September 2017, whereby they have been allowing bonds, including long-term Treasuries, to roll off their balance sheet. Thus if QE/Operation Twist had induced long-term yields to plummet, QT should have the opposite effect by driving the yield on 10yr Treasuries higher. While the 10yr yield did indeed start rallying higher from 2.05% in September 2017, it had peaked in October 2018 at about 3.23%, after which it has been plummeting (currently at around 2.60% at time of writing), even though the Fed has continued to engage in QT. While the Fed has signaled that it will end QT in September 2019 (and begin tapering in May 2019), the decline in the 10yr yield over the past several months even amid active QT from the Fed reflects serious deterioration in the outlook for future economic conditions, which would make it unreasonable to belittle the importance of a potential 2yr/10yr inversion. Therefore investors should not ignore the possibility of a recession after a 2yr/10yr inversion occurs, and should beware of poor S&P 500 performance amid the onset of a recession.

Bottom Line

While several investors are fretting over further inversion in the yield curve, historical data analysis suggests that the S&P 500 can continue climbing higher even as the yield curve flattens and inverts at the 2yr/10yr section. The Pearson’s coefficient correlation ‘R’ between the 2yr/10yr spread and the S&P 500 since December 2013 stands at -0.87, and results from my linear regression analysis offers significant statistical evidence confirming the negative correlation between the two variables. Hence, even as the spread falls into negative territory, investors should not underestimate the ability of the S&P 500 to continue climbing higher.

Nevertheless, an inversion at the 2yr/10yr section still holds significance in signalling a recession ahead from my perspective, as I do not believe ultra-loose monetary policy actions from the Fed have distorted the yield curve spread to such abnormally low levels that an inversion becomes less convincing as a recession harbinger, compared to previous inversions. The 2yr/10yr spread has been falling from a peak of 2.66%, similar to levels seen prior to the previous two recessions (between 2.60% and 2.75%), and the 10yr yield continues to drop even as the Fed engages in QT. Thus, investors should not ignore the potential for the inversion to presage the next recession. Historical data shows that the S&P 500 mostly tends to perform poorly during recessions, which is undesirable for investors.

Therefore, while investors should not run from the S&P 500 while the 2yr/10yr spread is inverting, they should also be wary of the fact that the previous three recessions occurred within a year of the spread reaching its trough (in negative territory), and recessions do not bode well for equity market performance. Based solely from a yield curve inversion perspective, let the 2yr/10yr inversion run its course (and reach its trough) before booking profits on positions in the S&P 500, and then avoid equities until the next recession passes and stocks have declined meaningfully back lower. Please keep in consideration that this research solely assesses the relation of the S&P 500 performance with the 2yr/10yr spread, and that investors should combine this research with other fundamental factors such as sales & earnings growth rates and valuation metrics when making buying and selling decisions.