20 March 2018
A little clarification before we begin: at StashAway, our investment process is systematic and driven by economic and market data. In other words, forecasting the future is not part of our investment processes. But, our system looks at a lot of data, and we thought it would be interesting for you to see what the data says about the economy and markets. Readers should view the content of this article as a holistic interpretation of the signals we observe.
StashAway continues to see a global economy that is expanding steadily at moderate pace. This stance is based on growth and inflation data, and it is further supported by search analytics that confirm this picture. Interestingly, search volume on Google related to negative themes such as “crisis”, “recession” and “unemployment”, continues to register as low interest by historical standards.
As the market continues to grow, we must bear in mind that we've been in this expanding economic cycle for more than eight years. Further, there are assets that are overvalued, and this overvaluation can make such assets vulnerable to even a slight slowdown in economic growth, regardless of whether economic fundamentals are fine.
This overall situation leads us to two main portfolio management takeaways: 1) In a continued positive economic regime, stay invested in assets that correlate with growth, but buy those that are less overvalued. 2) Don’t forget to diversify. Add protective assets to your portfolios whenever both 1) valuations are depressed and 2) when their price momentum are improving.
2017 was a particularly strong year for growth-oriented assets. In particular, growth momentum has been strong, has consistently outpaced inflation, and was supportive for corporate earnings. Equities in the Asia ex-Japan region, US technology, and Emerging Markets were the best performers of 2017, having returned 42.1%, 33.7%, and 30.2%, respectively (Figure 1).
When things were going so well, it was easy to forget the value that a protective asset brings to a portfolio. Hearsay and popular belief that government bonds and gold would perform poorly when central banks hike interest rates added to the confusion. It turns out that nothing is further from the truth. Data that analyses the performance of gold in past rate cycles going as far back as 1971 have shown that precious metals tended to do better in hiking cycles. This continued to be the case in 2017 when gold delivered a solid 12.8% in total return.
But what about US government bonds? Shouldn’t they suffer when the US Federal Reserve raised interest rates three times in 2017? Wrong again. In fact, 1-3yr, 3-10yr, 10-20yr, and >20yr US government bonds have returned positively at 0.6%, 2%, 4.4%, and 9%, respectively while also providing hedges against a stock market decline.
The above lessons from 2017 taught us that objective data analysis should drive investment decisions, and that economic factors are the ultimate driver of returns. One should not be overreacting to news flow, hearsay, and rumours. We continue to maintain the same principle when looking ahead to 2018.
Source: StashAway, Bloomberg
Starting with the US, the pace of growth remains greater than that of inflation. Growth, as proxied by industrial production, has re-accelerated from 1.46% yoy in August to 3.35% yoy by November 2017, while inflation has only increased from 1.9% to 2.2% yoy in the same period. Overall confidence in the global economy, as inferred from leading indicators, such as search analytics, is also strong. As shown in Figure 2, according to Google Trends, the search interest(2) for negative economic themes such as “crisis”, “recession”, and “unemployment”, continues to register low interest by historical standard since 2004.
It is important to understand that economic factors drive search trends, and so we do recommend looking directly at economic factors. But, we still thought that the Google Trend data are interesting, and available to anybody with just a few clicks. Looking back to the the financial crisis of September 2008, economic momentum had proven to be ahead of search trends; as early as April 2008, a slowdown in US growth had already been observed, clearly signaling that things had changed. A few months later, in August 2008, an entire month ahead of the Lehman Shock, search trends around the words “crisis”, “recession”, and “unemployment”, validated the slowdown.
Looking ahead, it’s most important to monitor trends and momentum of economic factors when assessing growth-oriented assets, particularly because their valuations are increasingly expensive, and so they could be more vulnerable to any negative economic surprise in the future. We will also watch out for any uptick in negative search interests as additional validation.
Source: Google Trends
When the economy is in an expansion phase, there is less economic uncertainty, which in turn reduces risk aversion. In response, the volatility of the stock market declines. Conversely, when the economy contracts, uncertainty about the economy heightens, making investors risk averse. In response to higher risk aversion, stock market volatility rises.
The ups and downs of growth, as described above, are known as the business cycle. As can be observed in Figure 3A, the current cycle of economic expansion, and therefore decline in volatilities, started 8 years ago (January 2011), which is longer in duration than either of the last two cycles. The business cycle between July 1988 and August 1995 experienced 6 years and 11 months of economic expansion and a decline in stock market volatility, and the business cycle between November 2002 and January 2007 had a shorter duration of 4 years and 2 months.
There are vulnerabilities building in the markets, although we have been in a benign economic environment for a period of time that is longer than past business cycles. When stock market volatility is substantially lower than growth volatility (Figure 3B), a slight slowdown in the pace of growth or earnings could be sufficient to trigger larger price volatility.
It’s important to adopt a holistic approach when making asset allocation decisions for any given portfolio. At times, the valuation of an asset may not be attractive (i.e. expensive), but the asset is simultaneously well-positioned to benefit from the prevailing economic environment. In this situation, an asset tends to have positive price momentum. In today’s market, this is particularly true for growth-oriented assets.
On the other hand, an asset may be attractively priced (i.e. undervalued), but is poorly positioned for the existing or upcoming economic cycle. This asset tends to have negative momentum, and suffer from ”value trap”. As a result, we present the following asset class outlook based on both valuation and price momentum.
Among the class of assets that are well-positioned for the current economic environment, technology, consumer discretionary, and small-cap growth stocks in the US are relatively more attractive. They have similarly strong price momentum, while also trading at less-expensive valuations (Figure 4).
In international equities, Asia ex-Japan and Emerging Markets are more attractive than Europe, as they are less-expensive and have improving price momentum (light blue bars in Figure 4).
Undervalued assets, such as investment grade corporate bonds and convertible securities in the US, can be valuable, too. Although they lack positive price momentum, they are not suffering from it either. This makes their deep discount to economic fair values attractive and suitable for investors with long-term time horizon.
Source: StashAway’s ERAA® Model
In good economic times, markets tend to under-appreciate, and hence undervalue protective assets. However, portfolios will always need a certain amount of protective assets to achieve diversification.
As it currently stands, gold is attractive, as it is substantially undervalued compared to its economic fair value. Further, the price momentum of gold is improving (light blue bars in Figure 5).
Consumer staples and inflation-linked bonds in the US are also interesting, as they are even more undervalued than gold is. But due to their lack of momentum (grey bars), they are more suitable for investors with longer-term time horizons.
Source: StashAway’s ERAA Model
As we like to say, we’re not in the business of forecasting markets or economic conditions. Instead, we assess economic data as it comes in to make portfolio management decisions that cut through market noise and volatility. Ultimately, the economy drives asset classes’ performance in the medium to long term.
The economy is still doing very well, but even in times of growth, asset classes may show positive or negative valuation gaps (overvalued or undervalued). The latter can lead to augmented vulnerability to future economic surprises or opportunities. When an asset class is overvalued, a slight slowdown in the pace of growth, or earnings, is all that is required to trigger abrupt price swings, even if fundamentals are in fine. For example, this is why StashAway’s customers today do not hold any S&P 500 in their portfolio. Depending on the selected risk levels, they have generally seen their allocation to Gold and Consumer Staples increased from December 2017. More generally, even when times are good, portfolios still need sufficient protective assets.
While the data indicates that the economy is still experiencing growth, investors should continue to diversify portfolios in order to harness the maturing growth business cycle and protect against potential sudden economic downturns. Do this by continually evaluating both valuation and price momentum of an asset class that is part of, or could be part of, a diversified portfolio.
(1) Momentum is the rate of acceleration of an asset’s price. Here, we calculate price momentum of an asset with the popular MACD(12,26,9) specification in technical analysis.
(2) “Interest over time” is a number that represents search interest relative to the highest point on the chart for the given region and time. A value of 100 is the peak popularity for the term. A value of 50 means that the term is half as popular. Likewise, a score of 0 means the term was less than 1% as popular as the peak.