Categories

Pragmatic Capitalism

Practical Views on Money, Finance & Life

What is the Worst Case Scenario for Bonds?

« Back to Previous Page
0

Actually, the worst case scenario for bonds was that 1940-1980 period (or technically 1953-1980 once the Fed could set rates again). Inflation was 5.97%/year from 1940 to 1950, 3.85%/year from 1940 to 1960, 3.54%/year from 1940 to 1970 and 4.65%/year from 1940 to 1980; correspondingly, the risk-free 10-year Treasury returns were 2.28%/year from 1940 to 1950, 2.09%/year from 1940 to 1960, 2.36%/year from 1940 to 1970 and only up to 2.76%/year from 1940 to 1980.

And keep in mind those inflation rates were not manipulated downwards via hedonistic adjustment malarky and all that jazz for that only started in 1980. So no matter how you spin it, you had a very serious permanent loss in purchasing power if you couldn’t hold on longer than 40 years to benefit from the new secular bull after the end of the previous secular bear. You would not have been able to retire in this time frame and not run out of money (I can’t remember if it was nominal or real) at a 4% withdrawal rate unless you had no less than 50% equity coupled with no more than 50% bonds (of the total market, medium-style duration type). Most investors don’t need that kind of timing stress.

Closed
Marked as spam
Posted by MachineGhost
Posted on 06/02/2016 9:22 PM
196 views
Private answer

Yes, I mentioned the real loss in the article. You shouldn’t expect to maintain your purchasing when you own an aggregate bond portfolio. You own stocks to protect against inflation. And even a marginal position in stocks of 10% stocks and 90% bonds would have resulted in a positive real return over the period from 1940-1980.

The key point though is that you don’t use bonds to protect against inflation risk. You use them to hedge the equity piece in a portfolio.

Marked as spam
Cullen Roche Posted by Cullen Roche
Answered on 06/03/2016 12:14 AM
    Private answer

    Really, even 10%? That’s impressive. But they definitely have to be shorter duration bonds for that to work? Duration always seems to get overlooked a lot in bond discussions and investors don’t realize the durations were radically different historically. A 30-year Treasury bond had a duration as low as 7-years in the late 70’s, believe it or not, so there was no reason not to buy it.

    Do you believe there is an optimal duration length to shoot for for all times? I can’t shake the feeling there is some rational way to determine what should be the optimal duration to be buying at any given time.

    As you may know, I break “bonds” down further because corporate bonds, etc. have credit and economic risk just like stocks do, whereas Treasuries are risk-free hedgers, so I’m a purist over using a total market aggregate. If you were really worried about sovereign risk, you could always supplement with pure triple A corporates (well AA+ nowadays).

    Marked as spam
    Posted by MachineGhost
    Answered on 06/03/2016 1:58 AM
      Private answer

      I agree MG, it’s important not to think of bonds as one aggregate asset class — there’s too much difference, and resulting opportunities, that would be (is) missed looking at it that way.

      (as an aside, I also feel the generalization of hedge funds on this site suffers from the same.. there’s much opportunity missed when you group the 75% bad funds, and 20% only “good” funds, in with the 5% of outstanding funds, that are worth the fees)

      Anyhow, back to bonds, while folks can talk all day about the “risk” and “worst scenario” for bonds, my favorite fund vehicle for long interest rate duration, TMF, is +36% so far this year. Preferred Equity and Sub Debt – my favorite spot in the capital structure currently – has offered nice opportunities for some time to book 6-10%+ annual yield, bought at discounts to par with upside in both lower rates and higher rates (!) — this bc if we see higher rates, it should be predicated on an improvement in economic conditions, increasing the creditworthiness of the issuer, and thereby the security’s price; and in lower rates the price increases bc the security’s fixed rate is worth more, relatively speaking… Munis also offer a ton of value in individual securities… especially longer, down in credit and/or 0%cpn bonds.

      Floating coupons are not much discussed when “talking bonds”, and those can be used to pull back overall price risk (duration) in a portfolio, allowing to take on more said risk in other securities.

      For example, building a portfolio of floating (low *interest rate* risk) bond of lower credit quality (higher *credit* risk) trading at a discount (like a State Street Bank trust preferred trading at $85, offering +2% over LIBOR), along with high quality (low *credit* risk) longer maturity (high *interest rate* risk) non-call bonds (like 15yr hi-grade munis at 4-5% taxable equivalent) I might suggest. If rates go up, your floaters reprice to higher interest cashflows, if rates go down, your longer bonds see capital gains… in the meantime, you’re clipping a nice cashflow.

      Remember the other side of the coin from “interest rate risk” is “price appreciation potential”… no one really talks about that when they talk about duration — duration loses you money, but also can make you money… like in the last 30+ years… and I think for the next several years.

      lastly, Cullen I’m not sure you mean ‘convexity’ when you are talking about the cumulative returns over time of a constant duration fund…. Convexity is the 2nd derivative of the Price and Yield relationship… the first derivative, duration, is how much the price changes per unit of yield change… the second derivative, convexity, is how much the price/yield relationship changes per unit of yield change. It is generally used to describe optionality of a bond – like with callable bonds, or pre-paying mortgage securities, where the option is held by the ‘issuer’, and – generally speaking – will always work against the bond investor. A bond exhibiting ‘negative convexity’ generally means that the investor is holding a bond with a call option to the issuer, and that the bond will be called away when the investor would rather be holding the bond’s cashflow (ie rates have declined), or not called when the investor would rather have his principal back to re-invest (ie rates have increased).

      Marked as spam
      Posted by jswede
      Answered on 06/03/2016 2:53 PM
        Private answer

        MG,

        There’s no right duration for everyone. It’s totally dependent on the risk profile of the investor. And yes, I totally agree on separating bonds into types. As you know, I am very firm about T-Bonds and their role as the only true safe haven. They’re a totally different type of bond. Likewise, most corporate bonds are more like stocks than bonds because they actually react like stocks in a crisis. So you have to be more granular. I was generalizing in this post because it’s easier to explain the concept being discussed.

        Marked as spam
        Cullen Roche Posted by Cullen Roche
        Answered on 06/03/2016 3:43 PM
          Private answer

          Swede, the whole post was about the convex relationship between price and yield! Just look at the chart I showed….

          Also, I don’t mean to demean hedge funds. There are some good hedge funds, but the funds most people refer to, the ones that invest in public equities, are shit.

          Marked as spam
          Cullen Roche Posted by Cullen Roche
          Answered on 06/03/2016 3:46 PM
            Private answer

            Bond convexity is change in price per change in price yield relationship, or change in price per change in duration.

            You are showing cumulative return per unit of time. I agree your chart is convex, but it’s not in any way showing the concept of “bond convexity”. Time is not a direct variable in bond convexity.

            Marked as spam
            Posted by jswede
            Answered on 06/05/2016 1:28 PM
              Private answer

              All the post does is show that duration declines as yields increase meaning that bond prices become less sensitive to price changes as yields increase. This is the most basic description of convexity. I think maybe you’ve misunderstood something in the post.

              Marked as spam
              Cullen Roche Posted by Cullen Roche
              Answered on 06/05/2016 1:37 PM