Correlations Flashing Yellow Light

Submitted by Nick Colas of DataTrek Research

Asset price correlations remain at elevated levels, even as US stocks stage a mid-quarter rally. Why? We attribute it to continuing macro concerns about everything from the price of oil to future Fed policy, geopolitics, and long-term interest rates. Bottom line: don’t let your guard down. Still high correlations + even a small market shock = much higher volatility.

At a naïve level, outperforming the S&P 500 has been easy in 2018: overweight Technology as much as you dare and snip your holdings of everything else in equal measure. Tech is up 9.2% on the year versus the market’s 1.8% advance. If you are a value investor, large cap Energy would have done the trick. After some hair-raising volatility in February the group is now up 6.2% in 2018.

In practical terms, there’s a big problem with that notionally easy recipe. The correlations between Tech and the S&P 500 have been +0.95 since February’s volatility. Yes, Tech’s strong fundamentals have helped tremendously (the group has been down on the year only 2 days since Jan 1). But in terms of portfolio diversification, it is a dud.

Large cap Energy’s ties to the S&P’s performance are somewhat looser, which when combined with the recent surge in commodity prices helps explain its YTD outperformance. Its price correlation to the S&P has been below the average industry sector for the last two months (0.71/0.76 vs. 0.75/0.80), so capital looking for less-correlated price momentum has elbowed its way into the group.

All this is important for one reason: correlations define the utility of capital markets for investors as much as simply making money in the market. “Beating” a rising market by taking on more risk is relatively simple. Managing returns in the context of risk and uncertainty requires access to investments that move differently from each other. The more options here, the better for both investors (who can put more capital to work) and the societies that uses their capital to grow and prosper.

US stocks and global capital markets have not delivered much on this point since 2008, price performance notwithstanding. Sector correlations in the S&P 500 were uniformly high from 2009 – 2016, as were the price relationships between foreign developed/emerging equity markets to US stocks. Last year saw the first real declines, which we attribute primarily to the 2016 US elections only because the data clearly shows that cause-and-effect.
This year has been a roller coaster for correlations as much as stock prices, with February’s volatility pushing up the former as much as they depressed the latter.

Here are three points on the most recent correlation data, calculated through yesterday’s close:

1. US sector correlations still remain much higher than 2017: Over the last 30 days, the price correlation for the average S&P sector to the market as a whole is 0.75. In 2017, it was 0.55. That’s the difference between an r-squared of 56% now to 30% last year. In other words, the daily move in the S&P 500 “explains” +50% of any sector’s one-day price action now versus just 30% last year.
2. Current correlation levels give a better picture of investor uncertainty than current VIX readings, which have declined of late: Correlations always spike during periods of market volatility such as earlier this year, but unlike the CBOE VIX Index they don’t typically fall quickly after a shock. Unwinding correlations takes longer because they more accurately reflect market uncertainty about longer-term macro factors like Fed interest rate policy, geopolitical concerns, financial stability, and long term interest rates. The VIX captures 30 days of expected volatility, after all, not 30 months.
3. It’s not just US sector correlations that remain high: International equities, both in developed and emerging economies, are still clustering in terms of daily price performance around the S&P 500. The correlation between the US equity market and the MSCI EAFE developed market index was 0.83 last month, down only modestly from the 0.92 – 0.95 readings of the last 3 months. MSCI Emerging Market equity correlations to the S&P 500 ran 0.80 last month, down from 0.88 – 0.91.

Bringing this analysis home to three investment insights:

1. Don’t be lulled into a sense of false security now that the VIX is back to 13 and everything feels more “Normal”. Still-high correlations mean even small shocks will quickly transmit through to equity markets. We remain positive on US stocks for 2018, but the correlation math is clear. This isn’t 2017, with industry groups charting their own path and smoothing out the market bumps.
2. Financials and Energy stocks are better positioned than most groups to show both less-correlated and positive returns. Consider two of the largest macro market concerns today: rising oil prices and higher long term interest rates. Should those eventually crimp economic growth, the best equity hedge will come through exposure to these groups. In our last reader survey (dated 3/30), Financials and Energy were 2 groups DataTrek readers favored most (along with Tech). We continue to agree with that sentiment.
3. For those with long investment horizons (5 years plus): the return of high correlations in 2018 after last year’s decline is an important signal. Most long run asset allocation models use decades of price correlations to arrive at recommendations for equity/fixed income/other portfolio weights. The last 10 years may look like an anomaly there, with high correlations caused by the Financial Crisis and its aftermath. What if, however, it is the new normal condition? If so, that would argue for greater allocations to alternative assets in order to achieve better diversification in the future.

Bottom line: stocks seem to have found their footing but this correlation data says we’re not out of the woods.