Ars Regendi Simulation Forum
Don't Be Scared of Rising Interest Rates - Printable Version

+- Ars Regendi Simulation Forum (
+-- Forum: English (/forumdisplay.php?fid=1321)
+--- Forum: Real economics (/forumdisplay.php?fid=3)
+--- Thread: Don't Be Scared of Rising Interest Rates (/showthread.php?tid=26912)

Don't Be Scared of Rising Interest Rates - Helsworth - 25.08.2014 21:05

Morningstar study shows that diversified portfolios perform just fine.

One of the most frequently asked questions that we have been hearing over the past few years has been, "What impact would rapidly rising interest rates have on our bond holdings?" We know we are not alone in hearing this concern, considering how often this topic is discussed in the media. As most investors are well aware, when interest rates rise, bond prices go down, so these ostensibly safe bond holdings now appear riskier than before.

These discussions often focus exclusively on how bonds will perform poorly in a rapidly rising interest-rate environment, and while that is an understandable concern, we think it is more valuable for investors to consider how multiple asset classes, such as stocks, bonds, and cash, might respond to rising rates. Most investors own these asset classes in their portfolios, and each asset class plays a particular role in helping the total portfolio achieve a goal. We think that narrowly focusing on bonds without considering how they interact with other investments can lead investors to make misguided decisions. Ultimately, the performance of the entire portfolio is what matters most to investors.
Accordingly, in this piece, we illustrate how a narrow focus on bonds can be misleading and then present a better solution for evaluating the impact of rapidly rising rates on a portfolio.

Wrong Focus
We came up with two approaches to answering investors' question regarding what might happen to their bond holdings in a rising interest-rate environment. Through both of these approaches, we will show how solely analyzing the bond results can mislead investors.
The first approach involves regressing interest rates against bond returns to arrive at a specific value for how much bonds could lose if interest rates rise by any potential amount. We evaluated historical six-month return periods, as this roughly reflects an investor's short-term time horizon. Our historical data cover rolling six-month periods from January 1926 through December 2013, which includes periods of falling and rising interest rates. Based on the regression results, we produced the first column of Exhibit 1, showing how an incremental rise in rates could have an impact on U.S. government bonds. For example, our results suggest that over a six-month period of time, a 1% increase in interest rates would hypothetically coincide with a 5.4% loss for government bonds.

We ran the same regression analysis for a few other asset classes to see how they would respond to changes in rates. We considered U.S. corporate bonds, U.S. high-yield bonds, cash, and U.S. stocks1. The results from these regressions are also displayed in Exhibit 1.
The findings from this study confirm several commonly understood principles. Corporate and high-yield bonds are likely to lose less than government bonds in a rising-rate environment. We also find that changes in interest rates do not influence cash returns, but rather the actual interest rate at the beginning of each period determines the cash return. Stocks appear to have a slight positive relationship with interest rates.

The regression analyses also provide a measure of how well interest rates explain the returns of each asset class, known as the R-squared value in statistical terms. We find that the beginning interest rate and the change in interest rate do a good job of explaining government-bond and cash returns, but they are less helpful at explaining corporate-bond and high-yield-bond returns. The results suggest interest rates do a poor job at explaining stock returns.

We believe R-squared should be a key part of the discussion on how interest rates influence bonds. A common way to reduce the interest-rate risk of a bond allocation is to add credit risk, such as high-yield bonds, or to add shorter-term bonds, with cash being the shortest-term option. The results in Exhibit 1 confirm that both high-yield bonds and cash are less sensitive to interest-rate changes than are government bonds. The R-squared measures, however, reveal that even though high-yield bonds and cash are less sensitive to interest-rate changes, there are other risks that these asset classes carry.
Our regression analysis did not differentiate between periods where interest rates rose sharply or slowly, so we developed a second study to take a look at how these asset classes would fare in different interest-rate environments.

For this second analysis, we used the same historical data divided into rolling six-month periods. We then calculated the average six-month returns of U.S. stocks, U.S. government bonds, U.S. corporate bonds, U.S. high-yield bonds, and cash during environments in which rates rose rapidly (0.5% to 2.5% in six months), rose slowly (0.1% to 0.5%), were stagnant (−0.01% to 0.01%), fell slowly (−0.5% to −0.1%), and fell rapidly (−3% to −0.5%). We further separated the data by the starting interest-rate level of the six-month period to show how these asset classes performed during times of low and high rates. This information is presented in Exhibit 2. The colors highlight the highest returning average six-month periods (green) and the lowest returning averages (red).
In the government bonds table, the red along the bottom row illustrates that when interest rates rise rapidly, government bonds have historically performed their worst.

Corporate-bond returns, as illustrated in the next table, have a similar history, although the asset class has performed slightly better than government bonds have in rising-rate environments. Meanwhile, the high-yield-bond data reveal that these bonds have posted significant short-term losses at times, but the losses have not been highly correlated with the interest-rate environment. Cash returns have historically been heavily influenced by the actual interest rate, but not by the change in rate. Finally, stock returns show a minimal relationship with interest-rate changes and levels.
In this second analysis, we again find that government bonds are sensitive to rising interest rates and have historically lost value when rates rose quickly. We also find, however, that the worst interest-rate environments for government bonds have not been as bad as the worst time periods for corporate and high-yield bonds. These other asset classes carry different risks, such as credit and liquidity risk, which are important when allocating across these asset classes.

Diversified Investors Need Not Fear Rising Rates
We believe that looking at how a multiasset portfolio performs in different rate environments better reflects the concerns of a typical investor than analyzing solely the bond portion of the portfolio. Therefore, we constructed three diversified target-risk models2 to see how they would perform in various environments. The models have equity allocations of 20% for the conservative model, 60% for the moderate model, and 80% for the aggressive model. The equity allocations include only U.S. stocks, while the bond allocations include U.S. government bonds, U.S. corporate bonds, U.S. high-yield bonds, and cash3. We repeated the second analysis as described previously, in which we constructed tables showing how these models would have performed in different interest-rate environments.
Exhibit 3 shows the results from the analysis. We find the returns of the conservative model to be particularly valuable; it suggests that a bond-heavy multiasset portfolio would have, on average, generated positive returns across all the historical interest-rate environments shown. By allocating a portion of the portfolio to equities and cash, both of which have performed well in periods of rapidly rising rates, the conservative model provided positive returns despite losses in its bond allocation. Meanwhile, the moderate and aggressive models had returns that were only loosely related to changes in interest rates.
Setting Expectations
We think this method of analyzing the impact of interest-rate changes on multiasset portfolios is an effective way to set expectations. Our analysis suggests that for well-diversified investors, fears of a portfolio meltdown are misplaced. While we focused on interest-rate changes for this study, the framework described can be used to evaluate how sensitive a portfolio is to other market factors, such as valuation, inflation, and economic growth.

RE: Don't Be Scared of Rising Interest Rates - Alexei B.Miller - 26.08.2014 04:55

How is this news?