Insight

Applied philosophy: Back to the 90s?

Applied philosophy: Back to the 90s?

Nostalgia for the 1990s seems to be increasing not only in fashion, but also in hopes for returning to its higher growth rates and more stable economic environment. Assuming the 90s will be back, what can we learn from US asset returns from that decade? Based on valuations, equities are unlikely to repeat their 1990s performance. We find US corporate fixed income more attractive, especially if spreads versus Treasuries behave in the same way as in the 90s. 

The 1990s are back! Undoubtedly the best decade for music, as far as I am concerned (think drum & bass, not boybands). Maybe not so great for fashion, but it is not for me to judge if we want to return to platform shoes or overalls (although who can resist rocking a pair of baggy cargo pants). For “old millennials” like me, being nostalgic about that decade is not just a spectator sport. Growing up in the ‘90s combined the best of the analogue and digital worlds with a unique taste of freedom especially for those shaking off the yoke of communism (for example, in Budapest, where I grew up).  

It was not too shabby a decade for financial markets, either. Using US benchmarks for comparison (they have the longest histories), it was the best decade for equities with 19% annualised returns from 1 January 1990 to 31 December 1999 (based on the MSCI USA index from the beginning of 1970). Of course, it helped that the end of the decade was also the peak of a stock market bubble. It was also the best decade for REITs with 14% annualised returns (based on the FTSE  EPRA/NAREIT index from the beginning of 1990), although the lack of earlier data makes this comparison less valid.  

US fixed income returns may not have matched those of the 1980s for obvious reasons (rates declined from historical highs after the Volcker-led Federal Reserve prioritised fighting inflation). However, they were still better than anything that came after, with annualised returns on government bonds, investment grade corporates and high yield at 7%, 8% and 11% respectively (based on ICE-BofA US bond indices). 

What made the 1990s such a fertile environment for investment returns? First, it had strong economic growth at just above 3% on average, admittedly below the near 4.5% growth rates seen in the 1950s and 1960s but well above anything that came after (based on year-on-year growth of quarterly real GDP). Second, as shown in Figure 1, apart from the 2010s when it was coupled with lower average growth, GDP growth rates were also the least volatile (“The Great Moderation”). Third, inflation fell from the heights of the early 1980s to something closer to the now well-established 2% rate, although staying above it for most of the decade. This allowed for relatively stable monetary policy, especially in the second half of the decade. The Fed reached its peak rate of 6% for that cycle in early-1995, not too far from its current level. After a pause and three 25 basis-point cuts, the Fed’s target rate stayed at 5.25-5.5% until September 1998. The rate cuts that followed afterwards were triggered by turbulence caused by Russia’s default and the collapse of LTCM. Finally, no major financial crisis occurred in the decade unlike in the 1980s (Savings & Loan crisis) and the 2000s (Global Financial Crisis). 

Figure 1 – United States average real GDP growth and standard deviation of growth by decade
Figure 1 – United States average real GDP growth and standard deviation of growth by decade

Notes: Past performance is no guarantee of future results. Data as of 30 November 2023. We calculate average US real GDP growth by decade and standard deviation using year-on-year change in quarterly data. Each decade starts in the first quarter of year 0 (1950 for example) and ends in the fourth quarter of year 9 (1959 for example). The 2020s include data from Q1 2020 to Q3 2023. 

Source: LSEG Datastream and Invesco Global Market Strategy Office 

As much as we would wish for the return of the economic performance of that era, there are significant differences. Debt levels are much higher (government, corporate and household), while the US government is likely to run a much higher budget deficit adding to that debt pile. Although rate futures seem to indicate a limited number of rate cuts as in 1995-96, a more rapid pace is expected (about 125 basis points to the end of 2024 as of 1 December 2023). The Fed’s “dot plot” suggests more gradual easing, although they are also applying Quantitative Tightening at the same time, which was not part of their toolkit 30 years ago. 

IMF forecasts are also indicating lower US growth rates of around 2% on average between 2024 and 2028, which looks more like the 2000s and 2010s than the 1990s. On the other hand, inflation is forecast to be closer to the 1990s at 2.3% on average in the same period. If the US economy were to grow at a rate close to the 90s, we believe that generative artificial intelligence would have to play a role (in boosting productivity). It is also important, in our opinion, how economies adjust to the investment required to achieve net-zero carbon emissions without a significant reduction in consumption. We tend to take long term forecasts with a pinch of salt, but a return to the conditions of the 1990s would be a better outcome than our base case scenario. 

Let’s imagine that the US economy overcomes these obstacles. How would we invest if the 1990s came back? First, valuations suggest we should Underweight US equities. Using the Datastream US Total Market index, the 12-month trailing dividend yield currently stands at 1.5%, 46% lower than its long-term average. In the 90s, yields did not fall below that level until the final 2 years of the decade, which at that time was considered excessive. 

Unfortunately, our dividend yield series for US real estate starts in 2001, so a comparison with the 90s is impossible (although we were Overweight the asset class in our latest The Big Picture – see Figure 5), while commodities did not produce extraordinary returns during that decade, so may not be remembered so fondly (using the GSCI Commodity Total Return Index, their best decade was in the 1970s). 

This leaves fixed income, whose rise in yields have prompted enthusiastic declarations that bonds are back! Yields on government and corporate bonds may not be high as in the 1990s, except for US high yield, for which yields are close to levels seen during most of that decade. However, we are in no doubt that these yields are as attractive as they have been since before the GFC relative to equities, for example, especially when it comes to US Treasuries and investment grade corporate debt. 

Treasuries look attractive to us at these levels (especially compared to the post-GFC period) but there may be better returns on other assets if our assumption of an economic recovery in the second half of 2024 proves correct. If the US experiences either a deep recession (positive for Treasuries), or if inflation meaningfully reaccelerates following a commodity supply shock for example (negative for Treasuries), we would be back talking about the 1970s or 1980s for asset outcomes (perhaps also implying very different fashion choices). 

For us, choosing between government bonds and credit instruments comes down to credit spreads versus Treasuries. That also highlights another phenomenon that has not happened since the 1990s. As Figure 2 shows, there has recently been a decoupling between weakening cyclical indicators, such as the ISM manufacturing PMI index, and investment grade spreads, which are priced close to long-term averages after tightening in the last 12 months (high yield spreads are tighter than average). The risk is that, in the short term, if US economic growth slows, these spreads may widen, although we think that would be at least partly offset (in absolute return terms) by falling Treasury yields.  

Nevertheless, our assumption that the US economy will reaccelerate in the second half of 2024 may dampen the rise in spreads as inflation stabilises. A looser monetary policy setting also implies strong returns for corporate bonds. We remain Overweight both IG and HY in our model asset allocation. 

Figure 2 – US ISM manufacturing PMI index and corporate investment grade spreads since 1984
Figure 2 – US ISM manufacturing PMI index and corporate investment grade spreads since 1984

Notes: Data as of 30 November 2023. Past performance is no guarantee of future results. We use monthly data from June 1984 to November 2023. ISM = Institute for Supply Management. PMI = Purchasing Managers Index. IG = investment grade. Investment grade spreads are calculated by deducting the redemption yield of the Datastream US Benchmark Treasury Index from the redemption yield of the ICE-BofA US Corporate Bond Index. 

Source: LSEG Datastream and Invesco Global Market Strategy Office 

Related articles