A year after the outbreak of the COVID-19 pandemic, global capital markets are still feeling its effects and following a familiar pattern.
About a year after the outbreak of the COVID-19 pandemic, investors are facing an unusual but not unknown situation. Looking to the equity market, the example of Tesla stands out: Tesla is worth more than the seven next largest car manufacturers put together. Looking at standard figures such as sales or profitability, this valuation is difficult to understand in economic terms. Similarly, Apple’s announcement that its plans to develop its own electric car can be considered another indication of an expected disruption of the automotive industry. It seems that some tech firms might carry-over their technological edge into all areas of everyday life. If you trust the expectations of market participants regarding Apple, the opportunities are boundless. Its valuation is currently about twice as high, measured using the price-earnings (P/E) ratio as the average for the last few years.
Other parameters, too, indicate a remarkable situation on the markets, with the grossly exaggerated value of individual stocks such as “GameStop” reminiscent of the excesses of the “New Economy” a good 20 years ago. M&A activities, which were essentially a wasteland at the start of 2020, have since reached levels last seen in 2007/08. Favorable financing conditions, as well as profound changes to the economic environment, form the basis of this takeover boom. Likewise, the high number of IPOs makes for a promising environment for companies. Rarely has capital – both equity and debt – been so affordable.
The late 1990s were also a time of similar growth euphoria. New, internet-based business models flooded equities markets and fueled high-hopes of future growth and booming sales at numerous companies – until the dotcom bubble finally burst in 2000.
Chart 1 shows how the Nasdaq 100 index performed through the three years around the dotcom bubble of 2000, with the current Nasdaq 100 performance overlaid on top. Over the last two years, the technology index has grown in value by over 100%. Whilst returns in 1999 and 2000 were lower at just over 40%, they were still well above the long-term average. Coming into the second half of 1999, the index experienced a dramatic rise, peaking at the end of the first quarter of 2000 after soaring by over 150% between January 1999 and April 2000.
Looking at the prevailing economic environment and the underlying mechanics, the two periods are certainly comparable. Firstly, share prices at the time rose swiftly, as today. Secondly, investors focused on growth stocks in both periods. Whilst the Nasdaq accelerated, the broader S&P 500 US market index rose by just 10% in the period 1999/2000, and by 60% in the period 2019/2020.
As the current market environment is subject to various forces, it is difficult to understand exactly what is truly driving it. One strength of economic models is that they reduce the highly complex nature of the real world and so can shed light on blurry fundamental situations and identify the key factors. This kind of asset pricing model has been in use since 1998 in the form of GLOCAP, which we train – like an X-ray machine – on the markets.
Chart 2 shows the sensitivities of the four state variables used by GLOCAP. The sensitivities are estimated by the model on a daily basis and describe the relationship between changes in state variables and equity returns. The chart highlights the dividend yield, which reflects the fundamental attractiveness of equities based on valuation, as well as the credit spread, which serves as an indicator of companies’ refinancing environment. While the model uses all four state variables to determine the optimal equity allocation, including both the term spread and the TED spread, this analysis focuses on the dividend yield and the credit spread. These determine how attractive an investment in equity is from a microeconomic perspective, and are currently the main drivers of GLOCAP’s overweight in equities.
It becomes evident that the sensitivity of the two variables – and thus the cause-and-effect relationship – is economically “wrong” at present and runs counter to economic intuition. A low (high) dividend yield generally reflects an expensive (cheap) valuation and should therefore have a negative (positive) impact on equity allocation. Accordingly, the low dividend yield at present – it is similar to in the second half of the 1990s – should put pressure on the equity allocation. Yet the negative sensitivity means it is having a positive effect.1 This raises the question of why GLOCAP is currently inverting the relationship between dividend yield and equity contribution. This phenomenon requires closer investigation.
History shows that sensitivities change sign when there is a change in economic logic and familiar correlations (e.g. between high dividend yields and a future rise in share prices) no longer apply. This is the case if high returns on the equity market today come up against a high valuation (i.e. a low dividend yield). As chart 3 indicates, a change of sign in both the dividend yield and the credit spread also became apparent in late 1998, early 1999. Sharp price increases that cannot be explained by the fundamentals occur if a majority of market participants expect inevitable growth.
The current equity market valuation means that market participants are expecting strong profit growth as well as a high returns in the future. Investors pay for this by foregoing dividends in the present. This causes share prices to increase considerably without the corresponding dividend payments that would otherwise be expected – and the long-standing link between the two values is temporarily broken. In these situations, the model places more emphasis on growth, which is reflected in a reversal of dividend yield sensitivity.
Historical data also suggests that an equity market in a growth phase is typically associated with a very expensive valuation (i.e. low dividend yield). There is empirical evidence that periods with strong growth expectations are frequently followed by below-average equity market returns (US in 1920s, Japan 1980s, dotcom in 2000, emerging markets 2003-07). Nonetheless, the dotcom bubble also demonstrates that high to very high returns can be generated in the short term as a result of high growth expectations. If, however, the expected growth rates are then lowered, this has a significant negative impact on equity prices.
Similar to the dividend yield, the sensitivity of the credit spread also currently contradicts economic intuition. The low credit spread in the market at present reflects a good refinancing environment, which should benefit equities markets and thus the equity allocation in the model. Counterintuitively, however, the sensitivity is positive, and so the low credit spread is straining the equity allocation. Looking back, this dynamic was also present during the dotcom period. At that time, too, the credit spread provided an indication of attractive refinancing conditions.
What can explain this? The combination of good growth prospects and cheap refinancing is attractive for companies, so there is no barrier to the financing of numerous projects. Companies thus take out more and more loans, but this increasing debt also increases the credit risk, just like the credit spread. Given the strong focus on growth and euphoria on markets, however, investors initially pay little attention to the credit risk. The result? A weakening or even inversion of the relationship between the equity market and the refinancing environment.
In addition, sensitivity to the credit spread becomes positive after a rapid and substantial increase in the credit risk. In this case, equity prices fall on account of a liquidity crisis. Ample additional liquidity is then supplied, causing spreads to narrow, propping up equity markets. In this case, the model sees the temporary high in the credit risk as a buy signal and increases the equity allocation by inverting the sensitivity, as seen in September 1990, October 1998, October 2011, and March 2020.
The positive sensitivity for the credit risk premium at present is likely to be shorter lived than that of the dotcom phase, as the market environment is not fully comparable. In March 2020, lending markets quickly stabilized thanks to decisive intervention by central banks and rigorous fiscal support following the severe turmoil on capital markets. The credit spread declined to a historic low. Given the abundant supply of liquidity, it can be assumed that the credit cycle will behave differently to that of the late 1990s. We can also assume that there will not be relevant credit defaults in the near future and that the high level of liquidity should keep credit risk premiums at their current low level.
Capital markets are currently in an environment that contradicts some conventional ways of thinking. The environment is determined by a charged relationship between fundamentals-driven and momentum-driven prices. Based on comparable past periods, it can be assumed that this may continue for some time – in the dotcom phase, it went on for around two years. Yet it is also clear that the situation will return to normal again, once today’s growth fantasies are reassessed and focus shifts back to potentially sobering profit figures. Systematic models can measure these developments on an ongoing basis and continually adapt the allocation to take account of this.
1 The state variables are standardised in the regression and so a below-average state variable becomes negative.