Market Musings 290325: Stock market ETF timing with the high yield spread

Market Musings 290325:
Stock market ETF timing with the high yield spread

This week, we are returning to the challenge that my friend Tony gave me. He is looking for a simple, fairly reliable system for investing for the long term without incurring high trading costs.

In designing an ETF portfolio system for Tony that can be easily followed month by month, we first need to set the objectives of any system that we're going to use consistently.

The biggest challenge for investors is typically themselves, and in particular their behavioural biases. Many people believe that they can time the markets, but this is in reality very difficult to do.

1. The goal of the system is not to optimise performance necessarily, but more to keep in line, or generate performance in line with the major stock indices when they're going up, but To reduce risk at times when the stock markets are going down. Simply said, the idea is to capture the majority of uptrends when they occur and to avoid the bulk of down trends when they occur. The idea of this is then to allow the investor to stay invested, using this system rather than panic.

The problem is that once you sell in a falling market, it's very difficult psychologically to buy back in, especially when that market may have risen back above the level you sold it at. However, using a system, firstly, hopefully reduces the times when you sell, and then also makes it easier to buy back in because you're following the system rather than trying to use your own judgement

2. It has to be easy to use.

3. It should use indicators that are easily available and don't take long to check.

The choice of indicators can be either economic or financial market-based indicators. Economic indicators seem seductive, because in general, stock bear markets occur around economic downturns, typically around economic recessions when the economy is contracting (shrinking) rather than growing.

4. We should not see too many transactions in a 12-month period, because we want to keep the costs of running such a system low.

Economic indicators: a timing issue

The difficulty with using economic indicators to time the stock market is that the stock market itself is a predictive mechanism for the economy. That is, the stock market already tends to predict forthcoming recessions, so you need an indicator that is good enough to

a) avoid too many false signals, i.e telling you to sell when actually the best thing would be to stay invested, while also

b) giving an early warning signal that is early enough to allow you to avoid the bulk of any downtrend.

Complicating the matter is the timing of the release of any economic indicator, which could be weekly or more typically, monthly or quarterly.

For the purposes of a monthly system, quarterly data is useless because it's too infrequent. But even monthly data can be difficult to use because the data takes time to collect and be released and is not therefore available for a few weeks after the end of the month, which means that you're always working effectively with old data.

Prefer financial market indicators as they are more timely

But much better therefore, are financial market indicators. They tend to be timely. They can be calculated daily, and you therefore have access at the end of the month to the end-month data reading, unlike with economic indicators, where you may have to wait longer for the data to be published.

But let's not forget, we don't want to make this too sophisticated, so we can't make use of any proprietary indicators that most people cannot have easy access to.

There's no point having an indicator that costs a fortune to have access to, because that then defeats the object of the exercise regarding keeping running costs of any ETF system low.

So when we look at financial market indicators, we're looking at financial marketing indicators that are easily accessible for free on the internet.

The best 2 stock market indicators that I have found

After studying a whole range of both economic and financial market indicators, the best two that I've come up with are,

1. An indicator based on the US U-3 unemployment rate, typically comparing the current unemployment rate with its own 12 month moving average. The idea of this indicator would be to flag a high risk of recession, which would be detected when the unemployment rate is itself rising and also rising above its own 12-month moving average. For me, this is the best single economic indicator I've come across for timing the stock market.

2. However, what I find better is a financial market indicator that is the US high yield credit spread. Now the credit spread, as I explained last week, is the extra yield that an investor demands to buy US high yield corporate bonds instead of equivalent maturity US government bonds. This spread, therefore is the compensation for the extra risk that you take, because clearly, if you buy high yield bonds, also called junk bonds, then you take  the risk is that these can default, i.e. either the coupons are not all paid, or you don't receive all of the initial investment back when the bond matures because the underlying company is in serious financial difficulty.

We have access to the US high yield spread, as it's called, going back to 1997 via the Fred website run by the St Louis Federal Reserve Bank in the US. Access is free and the data is updated on a timely basis.

Using the US high yield credit spread

So when does this US high yield spread indicator suggest that we should not be invested in stocks? The answer is, typically, when the high yield spread is rising, indicating rising risk of default, which in turn suggests a rising risk of economic recession.

The threshold that we look at typically would be 5%. Now the high yield spread doesn't often get as high as 5% but this has proven to be a good signal over the last 28 years.

So the signal to sell out of stock markets comes when this high yield spread is rising, and exceeds 5%.

When would you buy back into stocks? Once the high yield spread starts reliably falling from a high level (above 5%), then is the time to buy back into stocks.

Recall that below a 5% US high yield credit spread, you are always invested in stocks.

Remember, the idea here is to keep any drawdowns to a manageable level, as there is no perfect market timing system. What you consider a manageable drawdown is a personal choice, but I judge manageable to be somewhere around 20% drawdown from peak level, with a relatively quick time to recovery, so you don't wait years and years before returning to the former peak value of your investment portfolio.

The performance of such a system, using the high yield spread as an indicator for risk-on (i.e. buy stocks) or risk-off (i.e. avoid stocks and buy bonds or cash instead) has been as good or better than the underlying S&P 500 index in the US and the STOXX 600 Europe index in Europe.

In the 2 examples below, we are observing the US high yield spread monthly at the end of each month, and then following the rules above. The first chart has only 2 choices each month, either investing in the US S&P 500 index (including dividends), or in US dollar cash (at the prevailing cash deposit rate).

The second chart is the same, except that this time the choice is between investing in European stocks (the STOXX 600 Europe index including dividends) or a euro cash deposit.

For simplicity’s sake, I have not included trading costs, so in reality any performance would not be as good as shown in these charts, since there would be trading costs for each switch between stocks and cash, plus the annual management fee of the ETFs to consider.

The middle chart with red bars indicates the months when the high yield indicator flags risk off, and the system thus invests in cash rather than stocks.

Importantly, it has beaten the performance of buy-and-hold for both US and European stocks and while heavily reducing the drawdowns (the periods when stock markets are falling).

It has signalled risk-off (i.e. buy cash or bonds rather than stocks) about 20% of the time since 1994, notably avoiding much of the bear markets of 2000-2003 (TMT bust), 2007-09 (financial crisis), 2011-12 (eurozone sovereign bond crisis), and 2020 (COVID-19 pandemic).

However, it has also flagged up false positives in 1998 and 1999 during the TMT bubble phase when tech stocks were rising fast in a volatile fashion, as well as in 2010, and 2015-16. During these risk-off signals, it would have been better with hindsight to have remained invested in stocks.

This indicator has kept drawdowns to well under 20% that up to now. The only slight problem is that it is difficult to test this over a longer period than 31 years, because we don't have reliable access to a US high yield credit spread index before that, and that is largely because the high yield corporate bond market was not all that developed before the mid-1990s.

Improved performance version: Using physical gold for the risk-off asset rather than cash

For those who are able to embrace slightly more risk in their monthly investing system, we can improve the historical return of this system by using gold rather than cash in risk-off periods.

In the next article, I will consider the merits of using the 200-day moving average as a key indicator for a simple stock market trend-following system. To see how well that works in comparison to the system described here based on the high yield spread.

Have a good week

Edmund