Tuesday, October 27, 2020

Owning Today’s Long-Term Bonds is Crazy

Today’s long-term bonds pay such low interest rates that it makes no sense to own them.  There is virtually no upside, and rising interest rates loom on the downside.  Warren Buffett called this “return-free risk.”  He was right.  Here I explain the problem and address objections.

As I write this, 10-year Canadian government bonds pay 0.623% interest.  If you invest $10,000, you’ll get a total of only $623 in interest over the decade, and then you’ll get your $10,000 back.  This is crazy.  Even if inflation stays at just 2%, you’ll lose $1237 in purchasing power.

Even worse are 30-year Canadian government bonds that pay 1.224% as I write this.  Your $10,000 would get a total of $3672 in interest over 3 decades.  This is a pitiful amount of interest over a full generation.  At 2% inflation, you’ll lose $1738 in purchasing power.  Even a portfolio that only beats inflation by 2% per year would gain $8113 in purchasing power over 30 years.

All investments have risk, but there has to be some potential upside to justify the risk.  Where is the upside for long-term bonds?  The only upside comes if we have sustained deflation.  It’s crazy to risk so much just in case the prices of goods and services drop steadily for the next decade or three.

Some investors mistakenly think they can always sell bonds and collect accrued interest.  That’s not how it works.  With a 30-year bond, the government is promising to pay you the tiny interest payments and give you back your principal after 3 decades.  If you want out, you have to sell your bond to someone else who will accept these terms.  You don’t get accrued interest; you get whatever another investor is willing to pay.  Counting on selling a bond is hoping for a greater fool to bail you out.  If future investors demand higher interest rates on their bonds, your bond will sell at a significant capital loss.

If the interest rate on 30-year bonds goes up over time, that’s actually bad for current bond owners, because they have to live with their lower rate instead of receiving the new rate.  If 30-year bond interest rates go up by 1%, you immediately lose 30 years of 1% interest; you can’t just sell to avoid the loss because other investors wouldn’t happily take these losses for you.

Let’s go through some objections to this argument against owning today’s long-term bonds.

1. Stocks are risky.

It’s true that stocks are risky, but I’m not suggesting that investors replace long-term bonds with stocks.  Short-term bonds and high-interest saving accounts are safer alternatives.  A decision to avoid long-term bonds doesn’t have to include a change in your asset allocation between stocks and bonds.  For anyone willing to look beyond Canada’s big banks, it’s not hard to find high-interest savings accounts paying at least 1.5% and offering CDIC protection on deposits.  If long-term bond interest rates ever return to historical norms, it’s easy to move cash from a savings account back into bonds.  So, you don’t have to live with a measly 1.5% forever.

2. Investors need to diversify.


The benefit from diversifying comes from owning assets with similar expected returns that aren’t fully correlated.  However, the expected returns of today’s bonds are dismal.  We don’t really own bonds for diversification these days.  The real reason we own bonds is to blunt the risk of stocks.  It doesn’t make sense to try to reduce portfolio risk by buying risky long-term bonds.  Flushing away part of your portfolio with long-term bonds isn’t a reasonable form of diversification.  Short-term bonds and high-interest savings accounts do a fine job of reducing portfolio volatility without adding significant interest rate risk.

3. Long-Term bonds have higher interest rates than short-term bonds.

Historically, long-term bonds rates usually have been higher than short-term rates.  Today, however, high-interest savings accounts pay more interest than long-term government bonds.  But that’s not the only consideration.  Interest rates will change over the next 30 years.  If you own short-term bonds, your returns will change too.  However, if you buy 30-year bonds, your interest rate won’t change for 3 decades.  If interest rates rise, new short-term bond rates will be higher than your old 30-year rate.  Even if long-term bond rates rise, that won’t change your interest rate.  New investors will get more interest, but your bonds will still be locked into the same low rate.

4. All durations of bonds have done very well for almost 40 years.

Interest rates peaked in 1981 and have declined steadily since then (with some bumps along the way).  Investors who bought bonds that paid high interest rates have been able to sell them at a premium because their high interest payments keep looking better as interest rates on new bonds decline.  Unless you believe interest rates can go negative to -10% or lower, the next 40 years can’t look the same.  If interest rates ever return closer to historical norms, long-term bonds will get clobbered.  

5. My favourite ETFs contain long-term bonds.

Almost all balanced funds and bond funds contain some long-term bonds.  My favourite ETF company is Vanguard because of their focus on treating investors well.  Vanguard Canada has a great lineup of asset allocation ETFs that allow investors to buy just one ETF for their whole portfolio.  Unfortunately, these asset allocation ETFs, including the new retirement income ETF called VRIF, contain long-term bonds.  I’d like these products a lot better if the only bonds they held were short term.

Concluding Remarks

Throughout history, it has made a lot of sense to own a diversified set of stocks along with government bonds of various durations.  Looking back in time, we recognize stocks bubbles, but I’ve never been sure I was in a stock market bubble while it was happening.  So, I’ve maintained my allocation to stocks through thick and thin.  But it’s not hard to see the problem with long-term bonds today.  They have plenty of downside possibilities with almost no upside.

Unfortunately, almost all simple balanced investment options for Canadians include long-term bonds.  The simplest reasonable investing solution I see for a do-it-yourself Canadian is to own Vanguard Canada’s VEQT for the stock allocation, and place the fixed income allocation in short-term government bonds or high-interest savings accounts (not from a big bank).  For those using advisors, good luck convincing your advisor that long-term bonds aren’t worth owning.

None of this is intended as investment advice.  I spend time thinking and writing about how to invest my own money.  I’ve always preferred to keep my fixed income investments liquid, and I’m more certain than ever that I don’t want to own long-term bonds.

35 comments:

  1. Is it not that XEQT is more attractive than VEQT. Better yield, lower MER, better dividends??

    ReplyDelete
    Replies
    1. Hi Stanley,

      I think the differences are too tiny to worry about. I have no problem with choosing XEQT, but who knows how the MER and other features of these ETFs will evolve over time (most likely both will get slightly better). I always tend to lean towards Vanguard because of their long-term commitment to treating their investors well, but Blackrock is a fine firm too.

      Delete
  2. Wade Pfau suggests that those with no bequest motive should consider replacing their bond holdings with annuities. Annuity payout rates have held up fairly well despite low interest rates. This is because of the fixed mortality credits which are built in. Inflation risk is a problem though.

    Even better, I think we may see some tontine like products coming soon. Would love to see Vanguard step up and lead in this innovation.

    ReplyDelete
    Replies
    1. Hi Garth,

      In theory, replacing bonds with an annuity makes sense, but there are challenges in practice. In Canada, the annuity market is opaque (making it hard to know if you're getting a reasonable price), particularly for anything other than a non-indexed immediate annuity. The lack of indexing misleads people as to the long-term value of payments. With a fixed 2% annual increase in payments you have inflation risk, but without any indexing you have inflation certainty.

      I definitely like the idea of tontine-like products. Even better if the tontine is invested in stocks as well as bonds.

      Delete
    2. Another thing to consider is that annuity payout rates are based (in part) on long-term bond rates (in addition to mortality credits). So, while an annuity may be better than directly owning long-term bonds, that amounts to comparing the annuity to a terrible investment.

      Delete
  3. Deflation is a real risk though. Equities would incur significant damage. Long term bonds do provide diversification for this scenario.

    ReplyDelete
    Replies
    1. Hi Bhch,

      A little short-term deflation wouldn't do much to make long-term bonds look better. It would take sustained deflation for many years. In such a scenario, we'd have low wages and high unemployment, and the government wouldn't be able to tax us enough to pay back their bond debts. They'd have to keep printing money, which prevents long-term deflation. I don't see any plausible scenario where long-term bondholders do well.

      Delete
    2. Japan provided a plausible scenario. 1990s.

      Delete
    3. Hi Bhch,

      Japan's deflation in the 1990s wasn't enough to cause today's long-term bond rates to lead to reasonable returns. But if you believe that the world has a high probability of being headed for worse deflation than Japan's for longer than Japan's, then long-term bonds still look good to you. Of course, if that scenario doesn't play out, the long-term bonds will give terrible returns.

      Another thing to consider is that just holding cash for 30 years is almost as good as long-term bonds, with the bonus that the cash can be deployed into bonds of rates ever go up.

      I'll stick to cash and short-term bonds with my fixed-income investments.

      Delete
    4. BHCh made the following comment:

      "I have some in ZAG and IGOV. Mostly short term, cash or TIPS for my FI."

      Unfortunately, I've confirmed Google won't allow any editing of comments in Blogger (including the comment author's URL), and there appears to be no workaround for the problem that links to private profiles, such as BHCh's look like broken links to web crawlers. So, I have to keep deleting such comments.

      Delete
  4. Just watched at interview with David Rosenberg and Brian Belski and David like long term bonds (and gold). Go to 17:25. He thinks we've got a minimum of 5 years with lower interest rates. FYI. Good to read/see different opinions.+

    https://financialpost.com/investing/investing-pro/the-bear-and-the-bull-david-rosenberg-and-brian-belski-on-where-markets-are-headed-next

    ReplyDelete
    Replies
    1. Hi Larry,

      The problem is that even if they're right, long-term bond returns will be dismal. It's not good enough for interest rates and inflation to stay low. We'd have to have sustained deflation for many years for long-term bonds to give decent returns.

      Rosenberg suggests that long-term rates will get even worse, making current long-term bonds look better. But that amounts to hoping to sell bonds to a greater fool over the next 5 years. I don't like holding an asset with poor future prospects on the hope that someone will overpay by more than I did.

      Delete
  5. How about bond ETFs? Bond ETFs provides monthly or quaterly distribution which usually is higher than the inflation rate.

    ReplyDelete
    Replies
    1. Hi YGC,

      The only way this can continue is if interest rates keep dropping (into negative territory) or the funds include return of capital in their distributions.

      Delete
  6. By recommending that one should avoid long-term bonds, aren’t you taking an active view? I index both the equity and fixed income components of my portfolio under the assumption that both markets are (mostly) efficient. The current prices should reflect the aggregate market opinion. As such, long-term bond prices should incorporate market expectations on credit, inflation, interest rate, etc. (i.e. all the factors that influence bond prices). Your post seems to suggest that long-term bond are mispriced. If that’s the case, shouldn’t we all be shorting them?

    ReplyDelete
    Replies
    1. Anonymous,

      Essentially, I am taking an active view of the long-term return of long-term government bonds. I don't know what will happen in the short term for anything, and I don't know what will happen to a particular stock over any time frame. However, all Canadian bonds of a given duration pay exactly the same interest rate. If I buy a 30-year bond (or own one through a bond fund) I know exactly what the interest payments will be and exactly how much I will get back after 30 years. While it's possible that interest rates will go lower so that this bond will have a capital gain, this means that its future prospects would become even dimmer. It's clear that holding the bond for 30 years is very likely to give terrible results (unless we have sustained deflation for decades), so the only reason to hold it is in the hope of selling it to a greater fool. I prefer not to invest this way.

      Delete
    2. Anonymous,

      As for shorting, I avoid this because you can get harmed in the short term even if you're right over the long term.

      Delete
    3. OK, but you could currently make the same argument about equities. They are very expensive based on history and nearly all valuation metrics. This leaves you in the same same situation - you would need to sell to a greater fool to make money on stocks.

      Delete
    4. Anonymous,

      I don't think that follows. Stocks could easily average 1% below their long-term average returns for many years, eventually returning to more typical valuations. If this happens, stocks will still give better returns than other choices. In contrast, I know exactly what I will get if I hold a long-term government bond until it expires, and the return would be terrible.

      Delete
    5. BHCh left a comment saying that while we know the nominal return of long-term bonds, we don't know the real return. For long-term bond rates to give anything but terrible real returns, we'd have to have deflation worse that that experienced by Japan and for longer. Owning long-term bonds is a very high price to pay to guard against this possibility.

      Delete
  7. Thanks - always appreciated your perspective on investing.

    ReplyDelete
  8. That assumes that equity and fixed income investors are homogeneous. To a great extent, I think that's true of equity investors. Most equity investors want to maximize risk adjusted returns.

    But the fixed income market is different. There are major players in the fixed income market with different goals than you and me. There are institutions that want to match assets and liabilities, and are less concerned about maximizing risk adjusted returns. Examples would be pension funds and insurance companies (annuities).

    Also, don't forget the effect of the government. Governments have significant effects on fixed income markets, and their goals don't tend to be maximizing risk adjusted returns. And for the retail investor, governments have given them an advantaged in private markets ( HISAs, GICs), that institutional investors can't access. That's CDIC coverage, and credit union coverage provincially.

    Finally, I have read - although not seen data to support it - that fixed income markets are dominated by nontaxable investors. If you're in a high tax bracket, tax could take 50% of your fixed income return, versus 0% for a nontaxable investor. That makes a big difference, when it comes to investing.

    ReplyDelete
    Replies
    1. Anonymous,

      I agree that there are large institutional investors whose motives are different from yours and mine. Whatever I think about their willingness to buy long-term bonds at these prices, they are experts at what they do and aren't interested in my opinion. I'll stand by my argument on the value of long-term bonds to ordinary investors.

      I can believe that fixed income markets are dominated by nontaxable investors. It's richer people who tend to have taxable accounts, and I doubt many of them waste their money with long-term bonds. Further, the common advice to put bonds in your RRSP and stocks in taxable accounts (that I disagree with) is followed by many people.

      However, if I understand your final point correctly, you're saying that not paying 50% tax on bond income makes a big difference in tipping the scale in favour of owning bonds. I disagree. The income is so low that 50% of it isn't much. It's not tax rates that matter so much as it is the amount of tax you pay that matters.

      Delete
    2. I agree with your comments in the last paragraph.

      I'm presently reading "Foundations of Investment Management". There are $US15.6 trillion in US Treasuries. The single largest owner are global central banks at $US6.3 trillion. Global central banks may own a considerably smaller portion of other US bonds. However, Treasuries play an important role in pricing those other US bonds. So indirectly, global central banks can influence much of the US bond market.

      My point in the last paragraph is to point out why I don't believe in bond indexing. My goals are different than those of global central banks. My number one goal is to maximize risk adjusted return. But for a nonAmerican central bank owning US Treasuries, their number one goal may be to keep their country's exchange rate down.

      Global central banks are just one example of bone market investors, who have different goals than me. Another would be domestic central banks, with their quantitative easing. As mentioned previously, insurance companies and pension funds may put a higher priority on managing risk via bonds (hedging liabilities) than I would.

      Delete
    3. Anonymous,

      That's an interesting potential reason for why long-term bond prices make no sense for individuals. On a certain level, it makes no difference to my actions why long-term bonds prices seem crazy, but it's comforting to have an explanation why why the whole world isn't really crazy.

      Delete
  9. It might be more accurate to divide investors into tax sensitive versus tax insensitive, as opposed to taxable versus nontaxable. Foreign central banks will likely pay withholding tax on their Treasury bond coupons, but that may not be particularly important to them. Although I haven't seen data to back it up, my guess is that stock investors are more tax sensitive than bond investors.

    ReplyDelete
    Replies
    1. Anonymous,

      It may be true that central banks have different sensitivity to taxes than individuals do, but if we focus on individuals, long-term bonds have terrible returns, even when those returns are tax-free in a registered account.

      Delete
  10. If one accepts the hypothesis that investors, with goals other than maximizing risk adjusted return, play an important role in the bond market, then what is an investor with that goal to do? One strategy is to be selective about fixed income investing. That means security selection, aka active management. Here, GICs, HISAs and discount bonds can play a role. From what I understand, discount bond pricing doesn't take into account taxation; please correct me if I'm wrong.
    For some, maximizing CPP is a good idea. Finally, one can try to profit from it. If you think interest rates are lower than they should be, for someone whose goal is maximizing risk adjusted returns, you can short fixed income. An example is a 5 year mortgage, where ratehub gives the lowest rate at 1.54%. Interestingly, the only 25 year fixed rate mortgage is with RBC, and the rate is 8.75%. Margin loans and using a HELOC might be options, but that's really shorting cash.

    I'm trying to think of other ways that a retail investor could profit from the low rates on long term bonds, but I can't think of any others.

    ReplyDelete
    Replies
    1. Anonymous,

      Bond pricing, like the pricing of any other asset, indirectly takes taxes into account by the prices market participants are willing to pay (taking into account their own tax situations).

      As far as profiting from high long-term bond prices, the only ways I can see to do this involves leverage. Some people can reasonably handle a modest amount of leverage, but it's very easy to get into big short-term trouble with a strategy that appears sound over the long run.

      Delete
  11. This is a great post and great discussion. I find myself somewhat siding with "other Anonymous" who suggests that, by ruling out long-term bonds, you are actually taking an active stance. I use index ETFs and as such own LOTS of stuff, including companies I have never heard of, from countries I know nothing about. I own some Federal and Provincial and Corporate bonds of all different kinds, again about which I know little individually. I am sure, among all the above, are some "duds". And probably some duds about which, if I did some research, I could make a compelling argument that Country X or Company Y or Bond Z has almost no chance and should be avoided. But then, that's the whole point of passive investing, no?

    Also, don't try to fight the trend. Can't long-term bonds continue to do well if rates continue to drop? I ask this honestly, I know they are low now, but the finance community has been saying for basically the last 12 years, that I can remember, that "this can't go on much longer and rates have to go up". Yet the opposite has happened? One thing I have observed is that, while the reset may inevitably come, the trend can hold sometimes WAY longer than many think. And that's one of the reasons you don't try to be "active".

    ReplyDelete
    Replies
    1. Anonymous,

      I think active investing in government bonds is very different from active investing in stocks or corporate bonds. With government bonds, all bonds of a given duration have exactly the same return. With long-term government bonds, we know exactly what we'll get over those 30 years. Making a decision about whether to own them is very different from deciding which stocks to own.

      "Can't long-term bonds continue to do well if rates continue to drop?" Not by much more. If rates go too far negative, it would actually pay to get a large block of physical cash and store it like gold. This limits how far negative rates can go. It's true that so-called experts have been calling for interest rates to rise for a long time, but there is a hard limit somewhere slightly below 0%.

      Another thing to consider is that holding bonds in the hope that rates go lower is an active bet. I just look at what long-term bond returns would look like if I held them to maturity, and the answer isn't pretty. So some bond owners have to lose out somewhere along the way.

      I know it's unsettling to be reconsidering an investment plan that was supposed to last indefinitely. This was relatively easy for me because I've always thought it made sense to avoid long-term bonds for a different reason (I prefer to reduce stock-related risk by holding fixed-income investments that don't have the inflation risk of long-term bonds). But now I also have the reason that long-term bonds returns are so dismal.

      Delete
  12. As Michael is alluding to, diversification in government bonds isn't as important as in stocks or corporate bonds. And with government backstopping, HISAs and GICs become government issued securities. Also, there is the argument that with a bond index fund, bonds become sluggish stocks. In an index fund, bonds are no longer fixed income: you lose the certainty of return. However, if you want to keep a constant duration, that tends to be easier with index funds. However, the composition of bond index funds can change with time. At least in the US, bond index funds have quite a bit more Treasuries than they did 10 years ago. As a taxable investor, I can't make money in fixed income, and don't consider fixed income as an investment. However, it can be an excellent way to manage risk.

    ReplyDelete
  13. I couldn't resist adding this

    http://mrzepczynski.blogspot.com/2020/10/trend-following-strategy-as-new-safe.html

    "The Fed is now the largest holder of Treasury debt"

    ReplyDelete
  14. So interesting your article. Do you know the permanent portfolio?
    It has 25% of long term bonds for hedging equity. Not holded until maturity, like an +20y bonds ETF.
    Maybe the expected return of l.t. bonds is lower than stock and s.t. bonds, but even that, Harry Browne thought that it's worth for the negative correlation with stocks if there is an anual rebalancing of the portfolio.
    What do you think about it?
    In this article they talk about it too. They think that is better an active fund with low costs and lower duration that 20y.
    https://seekingalpha.com/article/4378268-3-risks-of-permanent-portfolio
    By the way, what do you think about holding gold and tips for inflation hedging?

    Kind regards.

    ReplyDelete
    Replies
    1. Anonymous,

      I am familiar with the permanent portfolio. I don't find it compelling. The 40-year bull market in bonds can't continue.

      I split my portfolio into the money I'll need in the next 5 years, and all the rest. Let's call these my short-term and long-term portfolios.

      I minimize the risk in my short-term portfolio. So, I hold cash, cash equivalents, and short-term bonds. These work fine for hedging equity risk.

      For the long-term portfolio, I don't hold anything I wouldn't be willing to hold for the long term. This excludes everything except stocks. Long-term bonds have dismal prospects. Short-term bonds aren't good for the long term. Gold's long-term performance has been dismal compared to stocks. I don't see the need to hedge equity for my long-term portfolio because time does a good job of that.

      As for inflation hedging, I'm satisfied with the inflation hedging I get from owning stocks. I might consider inflation-protected bonds like TIPS for my short-term portfolio, but not the long-term portfolio.

      Your mileage may vary.

      Delete