In the previous article, we developed the idea of the “miracle of least correlations” by adding the US treasury bonds to the portfolio of protective sectors. In this article, we will look at the fundamental reasons why the US treasury bonds work so well in conjunction with indices and why corporate securities are not a valid alternative to them.
Now brief information on the dynamics of interest rates
When the economy is doing well, and as a result, securities indices are also growing, the Central Bank increases the interest rate on loans issued in order to earn more money. This leads to a chain reaction of increasing interest rates for all current issues of debt instruments, such as shares, bonds, and deposits in banks. This means that past issues of debt instruments at a lower interest rate than now become less attractive, and their prices are falling.
When the economy is doing badly, and as a result, paper indices fall, the Central Bank lowers the interest rate in order to prevent a wave of bankruptcies associated with the excessive debt burden of enterprises in the crisis, when they are unable to pay debts at the previous interest rates. This means that past issues of debt instruments at a higher interest rate than now are becoming more attractive, and their prices are rising.
Let’s look at the chart of the dynamics of interest rates of the US Federal Reserve in comparison with the dynamics of the S&P500:
As you can see in Fig. 1, in the event of a recession and a reaction to it in the form of a rapid fall the S&P500 index, the Federal reserve begins to respond by lowering the interest rate in order to restructure loans and reduce credit pressure on businesses. And it really works, because the market rebound is quite rapid, after lowering the interest rate to the required level.
And now let’s see how the US treasury bond rate reacts to lowering and raising of the Federal Reserve’s key rate:
Rates on debt instruments are highly correlated with the behavior of the Federal reserve’s key rate (Fig. 2). I am sure that you have observed more than once that during the crisis the interest rates on deposits in banks fall. Roughly the same thing happens with all other instruments with a fixed interest rate.
Let’s now see how the package price of past issues of the US treasury bonds reacts to an increase or decrease in their interest rate:
Fig. 3: ETF TLT price (green) versus 10-year US treasury bonds’ rates (orange)
We see a very strong negative correlation (Fig. 3): rates have fallen and the value of bonds in circulation has increased. This is explained by a very simple example:
We bought a bond at a rate of 5% and repayment (payment on it) in 10 years. A year later, the rate was raised and now they issue bonds at a rate of 10%. Naturally, we want 10% annually, not 5%, so we need to sell the current bond in order to buy the one that is currently being issued at 10%. But no one wants to buy our bond with a yield of 5% from us, when now you can buy a bond with a yield of 10%, but not from us. Accordingly, we have to give a large discount on our bond. A discount of such a size that it would be pointless for us to sell our bond and we would lose as much as we would gain from the bond replacement operation. In other words, when the interest rate rises, our bond issued at the previous low rate becomes cheaper.
If the bond rate falls back to 5% next year, we will not have to give a discount on our bond at 5%, because the current bond issues that can be bought NOT from us are in no way superior in terms of the yield of our bond. But last year’s bond, issued at 10%, is bound to skyrocket in price. And since there are no more such high-yield bonds, no one is going to sell them cheap – only much more expensive than last year’s prices. In other words, when the interest rate falls, our bond issued at the previous high rate becomes more expensive.
This simplified example explains the negative correlation between the price of issued bonds and the current interest rate on their new issues.
Corporate bonds as an alternative to government bonds
In addition to the US treasury bonds, there are corporate bonds. They are a fixed return instrument as well. In theory, they should react to the increase or decrease in the interest rate in a similar way, but in practice this is far from the case. Now we will examine why corporate bonds are not suitable for our investment portfolio and how they differ from the US government bonds at the statistical and fundamental levels.
Why corporate bonds are not suitable for us:
Put simply, corporate incomes are not provided with permanent tax revenues, unlike the state. Also, in case of default, the state can simply print a large amount of money and pay off its debts through inflation. Corporations, however, do not have such an opportunity. Therefore, during recessions, non-government bonds have a high risk of default and subsequent non-payment of money on the bonds. All this leads to the fact that corporate bonds fall during the crisis, not giving us the necessary negative correlation with the market.
Let’s examine “junk”, unreliable bonds, as an example of an enhanced correlation with the S&P500:
Due to the increased risk of default that rises during the recession, unreliable bonds cannot provide us with a negative correlation with the S&P500 (Fig. 4, red box) in its falls (Fig. 4, black boxes). This type of bonds makes absolutely no sense for our portfolio, as it behaves almost identically to ordinary securities.
But what if we take high-quality corporate bonds?
Fig. 5: High-quality corporate bonds LQD (red) and SPY (blue)
Although the ETF of high-quality LQD bonds falls less and correlates less with the market, it still has a positive correlation with the S&P500 (Fig. 5, red box). No matter how reliable corporate bonds are, their risk of default still increases during a recession. They are also unable to provide us with a qualitative negative correlation with the market in deep drawdowns, which will stabilize our portfolio.
Now let’s see how the US treasury bonds react to the lowering of interest rates in market falls
A completely different thing! A high-quality instrument with a vanishingly low risk of default in a crisis, it responds to lowering interest rates in full force. This gives us a negative correlation with the market in its falls (Fig. 6, red box). This is exactly what we need to create a balanced and secure portfolio.
Let’s compare portfolios of different types of bonds and the S&P500 in a 50% to 50% ratio to understand the impact of each type of bond on the portfolio better:
As you can see, only the quality of the US treasury bonds gives us exactly what our portfolio needs – the lowest possible correlation with the market (Fig. 7, red box). The portfolio has the lowest standard deviation (Fig. 7, purple box), the lowest maximum drawdown (Fig. 7, blue box) and the highest yield (Fig. 7, green box)!
Let’s compare the performance of portfolios based on the eSharpe and CALMAR metrics that we already know from the previous articles:
HYG+SPY– 6.87 / 12.71 = 0.54 eSharpe and 6.87 / 37.64 = 0.18 CALMAR
LQD+SPY – 7.56 / 9.92 = 0.76 eSharpe and 7.56 / 27.55 = 0.27 CALMAR
TLT+SPY – 9.34 / 8.68 = 1.07 eSharpe and 9.34 / 17.45 = 0.53 CALMAR
The TLT + SPY portfolio [50%+50%] outperforms the lower-quality portfolios in terms of risks and returns by a wide margin, implementing the “miracle of least correlations” (hyperlink) using the lowest possible correlation of the US treasury bonds to the market.
Now you know exactly what bonds you need to include in your portfolio to maximize its efficiency and minimize risks.
All calculations are done in portfoliovisualizer. By clicking on the link, you can check all the results yourself: https://www.portfoliovisualizer.com/backtest-portfolio#analysisResults
Have a good day and see you in the following articles!