Stress Testing the Market’s Expectations for a Soft Economic Landing

Drew Beverly, Sales Specialist on the Quantitative and Risk Analytics team at FactSet, authored this article. 

Looking across U.S. equity and fixed income markets, investors are betting big the Fed will zig (cut rates) even though the central bank has repeatedly assured it will in fact zag (hold rates higher for longer).

The US economy is still moving, consumers are spending, the labor market is tight, and inflation is still a problem. Yet, the investors vs. Fed disconnect has created a rift of sorts, with asset prices largely reflecting investor expectations rather than Fed forecasts.

Using the data below, we’re working with clients to help them understand how their portfolios might shape up if this hiking cycle turns out like each of the ones before it. While this time may in fact turn out to be different, as the market has so clearly priced in, FactSet’s stress testing and risk management capabilities can help identify exactly how portfolios might perform should we see a reversion to historical trends.

The Bond Market Disconnect

If history is any indication of what’s to come, we should expect the Fed to maintain an elevated policy rate until material weakness has emerged—the likes of which we haven’t seen in this cycle yet. The Fed has not made a habit of cutting its target rate to preemptively address a contracting economy. Look no further than the past three rate hiking cycles to see this in practice. In 2001, the Fed funds target was held at its terminal rate for nearly eight months. In 2007, it remained there for nearly 15 months. And in 2019, it was there for seven months before the first cut.

Historically, it hasn’t been until weakness has materially trickled through to the Fed’s preferred indicators—rising unemployment rates resulting in weakened consumer demands and lower inflation, for example —that we’ve seen a pivot. Because of the lagging nature of these indicators, however, landing the plane softly, so to speak, has been nearly impossible for FOMC members in years past. Despite this and recession warnings coming straight from the Fed, investors are unequivocally betting that this time may be different, as evidenced within areas of the bond market.

Figure 1: Fed Funds Target (Blue) Charted Against 2-Year Treasury Yield (Green) And the Target Minus The 2-Year Yield (Purple)

Source: FactSet

Using the Fed funds target rate and the short end of the Treasury curve, we can see the clear disconnect between investors and the Fed. The Federal funds target rate sits at 5.25% while the 2-year yield sits below 4%—a massive spread, on a historical basis, considering the impact of the Fed funds rate on short-term interest rates. Adding the spread (purple in Figure 1 above) between the Fed funds target and 2-year Treasury yield, we reveal that a gap of this magnitude has only been seen twice over the last 25 years, with both occurences immediately preceding recessionary periods.

There’s a number of reasons this disconnect could exist, the most likely of which points to investors continuing to discount the Fed’s commitment to this hiking cycle. Based on Fed funds futures, the market has priced in three 25 basis point cuts this year (Figure 2) while the most recent comments from Jerome Powell and FOMC members have suggested holding rates higher until at least 2024 (Figure 3).

Figure 2: Probabilities for Future Fed Rate Hikes (as of May 10, 2023)

Source: FactSet

Figure 3: FOMC Members’ Target Levels for Fed Funds Rate

Source: March FOMC meeting

This leads to the all-important question: Is the Fed suggesting this time will be different? In other words, are Fed actions or commentary giving investors reason to believe a soft landing is attainable? In February, that certainly was the case. However, based on the Fed’s March meeting minutes, the central bank’s narrative changed significantly. Per these excerpts from March FOMC meeting minutes, the Fed has made it crystal clear their priority is to tame inflation, not sustain economic expansion:

• “With inflation remaining unexpectedly high, participants expected that a period of below-trend growth in real GDP would be needed to bring aggregate demand into better balance with aggregate supply and thereby reduce inflationary pressure.”

• “Maintaining a restrictive policy stance until inflation is clearly on a downward path toward 2% would be appropriate from a risk management perspective.”

Further weak growth is needed, and the higher-for-longer trajectory is very much still at play. Despite predicting a recession, FOMC members unanimously chose to further their tightening efforts. With inflation stubbornly remaining well above the Fed’s long-run target of 2%, upside inflation risks (i.e., inflation continuing to rise) was a focal point as the minutes stated “Participants generally saw risks to inflation as weighted to the upside, though they also recognized some downside risks to inflation.”

Despite this, the bond market is pricing in almost 200 basis points of Fed cuts over the next 18 months, a 37-year high, according to investment strategy firm Macro Compass.

But it’s not just the Treasury curve and Fed funds futures market that are signaling a Goldilocks scenario ahead. Corporate spreads have remained uncomfortably tight despite Fedspeak predicting a recession and key economic indicators flashing warning signs. Drawing attention to high yield, spreads are trading barely above 400 bps, below the 20-year average and far from the median recessionary episodes (1,000 bps). Investors aren’t yet demanding extra compensation for the added risk of investing in riskier assets like we’ve seen previously during tumultous times.

Figure 4: ICE BofA US High Yield Spread

Source: FactSet

As seen below, investment grade spreads paint the same picture.

Figure 5: ICE BoFA US Corporate Spread

Source: FactSet

Looking at U.S. equity markets, the story is largely the same. From the low on October 13, 2022, the S&P 500 has risen over 18% as prices have been pushed higher in anticipation of rate cuts. This, too, would prove to be a first if we don’t retest those lows. If October 13 is eventually confirmed as the market floor, it would be the first time the yield curve (3m10y) inverted after the market low.

Historically, yield curve inversions have preceded market lows by as much as two-plus years. This time, however, the yield curve inverted 12 days after the market low set on October 13 last year. Furthermore, market lows have historically come after the Fed already started cutting and after a recession (according to National Bureau of Economic Research standards) has already begun. All this to say that unless this time is different, history would suggest investors should prepare their portfolios for a sizable drawdown from here.

There are two ways we can view this disconnect. Either the market has gotten it wrong, or this time really is different. History would suggest one of two outcomes (detailed below), neither of which is currently priced in. With FactSet’s stress-testing capabilities, we can help you prepare for both outcomes.

Stress testing, whether through a pre-defined or customized test, allows the user to fully decompose risk measures at the broad portfolio level—or more granularly at the factor, group, or asset level. Using FactSet’s stress-testing application, we’ve forecasted two scenarios, detailed below, with a range of simple portfolio allocations.

Scenario No. 1

The first scenario is that the market has gotten it wrong, and short-end rates will need to rise as the Fed continues to hike and maintain elevated levels of its policy rate. With the Fed now forecasting a mild recession and prioritizing a hawkish policy stance to reel in inflation (rather than a dovish stance to support economic growth), we can also prepare for falling equity prices and widening spreads, which have historically accompanied economic contractions.

Stress Testing Inputs

• US 2-Year +100 bps: Shocking a single factor, the 2-year U.S. Treasury yield, up 1%

• US 10-Year +100 bps: Shocking a single factor, the 10-year U.S. Treasury yield, up 1%

• US 30-Year +100 bps: Shocking a single factor, the 30-year U.S. Treasury yield, up 1%

• High-Yield: Shocking a single factor, the ICE BofA US High Yield Index, up 2%

• S&P 500 -10%: Shocking a single factor, the S&P 500, down 10%

• Higher for longer: Combining the above individual shocks into one stress test

Figure 6: Scenario No. 1 Stress Testing Output

Source: FactSet

Scenario No. 2

The second possible outcome, on a historical basis, is that the market has gotten the direction right but the magnitude wrong. Historically, transitions to a dovish from a hawkish stance have been caused, more or less, by one thing: a contracting economy.

For lack of a better way to put it, the Fed has historically maintained an elevated terminal rate until something material in the economy has broken. If the market is right and cuts are coming, history would suggest we should expect swift cuts, the likes of which aren’t priced in. Prior to each of the last three recessions, the Fed did not simply cut rates a few basis points to stymie a contracting economy. Rather, Fed cuts were absolute and sharp.

• In 2001, the Fed cut its target rate 73% over the course of 343 days

• In 2008, it slashed it by 100% over 456 days

• And in 2020, it again reduced it by 100% in just 229 days

Each scenario has been unique in its own right, but the message remains the same: Each time we’ve seen the Fed cut rates over the past 25 years, it has been a clear and dramatic transition in policy. If you believe there will be three cuts before 2024, as the market has implied, history suggests you should prepare for the trio to be immediately followed by further substantial cuts.

As noted above, equity market lows historically have been set after an economic contraction has already begun, according to National Bureau of Economic Research standards. And that has been accompanied by widening credit spreads and short-term rates falling dramatically, among other things.

Stress Testing Inputs

• US 2-Year -250 bps: Shocking a single factor, the 2-year U.S. Treasury yield, down 2.5%

• VIX +50%: Shocking a single factor, the Volatility Index, up 50%

• High-Yield Shock: Shocking a single factor, the ICE BofA US High Yield Index, up 500 bps

• S&P 500 -20%: Shocking a single factor, the S&P 500, down 20%

• Aggressive rate cuts: Combining the above individual shocks into one stress test

Figure 7: Stress Testing Scenario No. 2 Output

Source: FactSet

With each shock varying in severity, we monitor how we might expect a handful of common strategies to perform under a range of scenarios. As expected, the higher-for-longer composite shock led, far and away, to the lowest expected return across each portfolio. To begin to understand why that’s the case, let’s take a look at how the results are generated. A preview of the historical periods that generated our stressed results is included below (only three days are shown for brevity).

Figure 8: Stress Testing Underlying Data Preview

Source: FactSet

As indicated by the blue text in the right side of the image, there are two methodologies to generate stressed portfolio results. Event-weighted and time-weighted stress-testing methodologies have identical calculation processes, with the difference between the two being the data fed into the tests.

The process involves calculating the covariance—and thereby correlations—between the stressed factor (the shocks in our higher-for-longer composite for example) and the factor returns of each factor in the risk model. With these calculated covariances, we can then find the beta between the stressed factor (or factors in our example) and each of the factors of the risk model.

For example, the beta between one of our stressed factors (e.g., S&P 500) and size factor might be 0.5. As a result, for each unit of exposure to size, every 1% change in the S&P 500 factor will have a positive 0.5% impact. So, if we were to shock the S&P 500 upward by 20%, the impact to the size factor would be an increase of 10%.

For further clarification:

• Event weighted: The model assigns higher weights to the periods with factor returns similar to the predefined factor shock amount. In our first stress testing example above, the model would assign higher weights to the periods with factor returns similar to the US 2-year Treasury yield up 100 bps, high-yield spreads up 200 bps, and S&P 500 down 10%.

• Time weighted: Here, the covariances weigh recent periods more heavily. As such, more recent periods would receive higher weights regardless of how similar they are to the shock amount.

For our event-weighted return columns in our example stress-testing output above, the model generating the stressed results has assigned the highest weights to the periods with factor returns similar to our defined shock amounts. We can see that in practice with our higher-for-longer example in Figure 7.

The three highest-weighted days used to generate our stressed returns are June 20, 2013, March 18, 2020, and June 13, 2022, all amid challenging market environments. Intuitively, this makes sense considering our higher-for-longer composite example includes dramatic shocks like the S&P 500 falling 10% and high-yield spreads widening 300 basis points, both of which are typical during times of market turbulence and instability.

Alternatively, we can monitor a time series of stress testing results to understand the changing impact of any given factor or scenario shock on the portfolio over time. As seen in Figure 8, the absolute risk stemming from our higher-for-longer composite shock has varied somewhat significantly, while the risk shown by our aggressive rate cuts shock has held more constant. We’ve also chosen to view these risks over time on a relative basis, as shown by our green and orange columns.

Figure 9: Stress Testing Time Series

Source: FactSet

To further identify portfolio vulnerabilities and potential risks, we can drill into each of these portfolios granularly—at the asset level or more broadly at a factor or sector level, for example—to help uncover pockets of risk that can be mitigated.

In Figures 10 and 11 below, we’ve shocked the yield curve higher by 100 bps, and we’ve grouped our portfolio by sector and sector subgroup. We then sorted our active return (event-weighted) against our benchmark, thereby allowing us to identify the portfolio’s largest risks.

Figure 10: Stress Testing Detail as of Oct. 31, 2022

Source: FactSet

Figure 11: Stress Testing Detail as of April 28, 2023

Source: FactSet

We can see that our active risk rose nearly 50 basis points in the six months preceding April 28, 2023, largely attributable to heightened risk from our financial sleeve. Alternatively, our industrials risk fell slightly relative to the benchmark over the same period. In this example, we’ve chosen to view the risk exposure differences over a six-month period. Depending on the insight sought, we can also choose to view these differences over monthly, weekly, or even daily timeframes.

Stress Testing Your Portfolios

Whether it’s recession risk, rate cut risk, or something else entirely, FactSet’s customizable stress testing capabilities allow managers to better understand the risk they may face under any set of circumstances. In this analysis, we stressed a number of portfolios relative to their respective benchmarks under individual shocks as well as composite shocks. To understand how your portfolios may perform within the confines of either scenario discussed here, please reach out to your FactSet MAC risk specialist.

Mr. Drew Beverly is a Sales Specialist on the Quantitative and Risk Analytics team at FactSet. In this role, he leverages FactSet’s Multi-Asset Class Risk solutions to help clients meet the unique portfolio construction challenges presented in today’s market environment. Mr. Beverly earned a Bachelor of Science in Finance from the University of Dayton.

This blog post is for informational purposes only. The information contained in this blog post is not legal, tax, or investment advice. FactSet does not endorse or recommend any investments and assumes no liability for any consequence relating directly or indirectly to any action or inaction taken based on the information contained in this article.