The Long View on US Government Bonds

One thing I constantly hear is “interest rates only have one direction to go – UP!”  There’s this myth that t-bond yields and interest rates in general just have to go higher.  But history does not prove this at all.  In fact, history tells quite a different story.

The chart below helps put things in perspective.  Since 1871 US Treasury Bond yields have averaged 4.3%.  Today’s rates of 2.8% are certainly lower than that, but not at record lows.  We’re still about 1% off those levels seen at several points in the past 125 years.

It’s also interesting to note that the high rates of the 70’s are a substantial anomaly in the data. It looks like many are suffering from a case of recency bias here.  And by recent, I do mean the 70’s.  That’s not entirely inappropriate given the long duration of these bonds, but when one steps back and reviews the true long-term history of bond yields the current environment looks much more benign than most imply.

(Chart via Hoisington)


Got a comment or question about this post? Feel free to use the Ask Cullen section, leave a comment in the forum or send me a message on Twitter.

Cullen Roche

Mr. Roche is the Founder of Orcam Financial Group, LLC. Orcam is a financial services firm offering research, private advisory, institutional consulting and educational services.

More Posts - Website

Follow Me:


  1. You are showing us a long term chart of nominal interest rates, and you are telling us that the high interest rates of the 1970s were an anomaly. In order to justify your assertions, are you also going to demonstrate to us that the inflation rates have been constant since 1871?

    Or, alternatively, will you show us a long term chart of real interest rates and go back to square one with your ‘analysis’.

    Kind regards,

    PS: By the way, can you point me to the raw data that have been used to produce the chart shown?

  2. from 1920 to the late 1940s price held stable. since then there´s a consistent trend up.

    i agree that the 70s price shock is unlikely to return unless we see a mideast war.
    but when speaking of anomalies you would then have to back out at least those stable 20+ years from the calculation.
    the current system ís designed for mild inflation. stable prices, below 2% inflation will never be tolerated again and a 30s hyperdeflation is unlikely as well.

    i couldn´t find inflation data before 1912. but this was likely a period of stable prices to (no major wars).

  3. Excellent chart. The spike in the 70’s was definitely an anomaly for at least two reasons. First, going ff the gold standard caused some initial volatility and uncertainty. Secondly, Paul Volker decided to let interest rates fly when he targeted money supply instead. That unfortunate experiment will not be tried again!

    But I’m not sure the yield chart prior to 1913 is too useful given there was no Fed.

  4. The T-bond bears who have advocated buying TBT (inverse T-bonds) have been getting scorched repeatedly for the past 4 years.

  5. I’ve seen a chart of prices from 1800 on. Sure wish I could find it. (from Siegel, maybe?) Prices varied by 2:1 over that period, although the long-term average was indeed fairly stable. Recall there were multiple panics during that period and the Civil War, pushing prices all over the place, despite the gold standard.

  6. Nice chart. To get a clearer perspective though, some more datapoints need to be added – Debt to GDP, government spending as % of GDP (discretionary and non-discreationary), and taxes as % of GDP. Of cause, real rates too, as victor rightfully mentioned.
    I think that 2% is achievable now. All the above data is just neccessary to understand what will happen to us when rates start to go up…..

  7. KB, if you are worried about the impact of rising rates on the so-called “creditworthiness” of the US govt, dont’t worry. It will have not impact.

    It might even be positive for the private sector to the extent that rising rates cause a higher govt deficit, which would be stimulative. Though it would make a difference whether you were a borrower or a saver.

    Of course, if rates rise, bond holders would be f-cked. But that would only be a paper loss. If they hold to maturity, they should be OK.

  8. when the Fed announces that rates will be going up – presumably at least a year or 3 from now – TBT will be in everyone’s cross hairs.

  9. Really, you think the economy will be strong enough the Fed will raise interest rates in 3 years time? I’ll take the other rise of that trade.

  10. I am not worried about “creditworthiness”. Let the government worry about such stuff. Also, I am not worried about “private sector”. I am worried about my trading position.

    Regarding so-called “paper loss” – are you serious??? Are you going to keep your 30Ys to maturity??? Do you actually hold them? Because I do.
    For your fun, find a simple excel model, and try to check where 30Ys would go if rate is up to “average” 4.3% or to 6%. Also, where do you think rates can go if they start to go up? Then, try to hold this stuff to maturity to avoid “paper loss”. What is your age, by the way?
    And then, if you are 20 years old now, and hopefully get your 30Ys matured in due time, what do you think would be the purchasing power of USD you’ll get?

    These are tough questions. The fact I am long US bonds does not mean I will be holding them to maturity and tolerate any “paper loss”. I am not a clown or “economist” for that matter….

  11. Depends on how long the household balance sheet recession lasts. There has certainly been improvement in that area.

  12. Didn’t know where you were coming from KB. Thanks for the clarification. The 30yr T-bond is for kids. Try putting the 30yr zero into your 6% model :)

  13. I don’t think most quality analysis of the “bond bubble” focuses on nominal yield, so that may be a bit of a straw man. The argument seems to be based around the collapse of real interest rates, which at approximately .5% are, even by the standards of the above chart, far below their long-term averages. That does not mean they are destined to immediately revert. But they should be recognized for what they are; clearly abnormal and risky for those relying on them.

  14. I believe the natural state of the economy was deflationary until the early 1900’s – about the time the Fed was created. Since then it’s been the governments policy to insure there is a constant state of inflation, at least to the degree they don’t blow up the economy.

  15. Obviously it was. No new money and expanding economy/population = deflationary. Except never smoothly, always through boom & bust.

    I wish goldbugs would understand that instead of claiming everything is smooth sailing with a gold standard.

  16. I know, I know…. I thought about buying those, but went to coupon paying – I wanted the yield!!

  17. I hear you, KB. You might want to consider a high quality 30yr corporate bond, which will give you a similar duration kick as the T-bond but with a higher yield. It should also give you better protection if yields rise as the spread is likely to narrow in a rising yield environment. Of course, the reverse is also true, i.e. the spread will probably widen if treasury yields decline. It’s a bit of a tradeoff.

    Please note that the above is not meant to be investment advice. Invest at your own risk etc, etc.

  18. P.S. If you already own the T-bond, you should be able to swap it directly for a corp, thus avoiding any bid/offer spread.

  19. Thank you, I understand the balance sheet recession and its implications. And I know that the household’s flow (interest payments as a percentage of real disposable income has improved) of debt has improved. What I really wanted to know is what data hangemhi is looking at that would cause him to even suggest interest rates will rise in 3 years time, let alone 1 year.

  20. Two other charts
    1. Long term rates from 1790 till now. It seems rates were more sporadic then now.

    2. Real interest rates from 50s onward.

    2. Dutch yields from the 1500 shows a few spikes aside from the 70s but what looks like a steady decline to 0.

  21. The 1790 chart is interesting in that the bond yield appears to have moved in a cycle of approximately 40 years. There were 4 lows in 1824, 1860, 1900 and 1946. Adding 40 years to 1946 should have given us a low around 1986. If there was one then, it must have been relatively minor because it is tough to see anything in the chart. In any case, adding another 40 years would put the next low in 2026. Seems a little far out there but who knows?

  22. But if rates drop further, I’ll take a non-callable bond any day. I’m not saying I like 30 year t-bond rates, but non-callability is a huge trait.

  23. I would also like to add that Japan, the poster child of the “rates can always go lower” camp, is currently issuing 30-year debt at a real interest rate of +2.2%. Again, this is far greater than the current real 30-year rates in the US of +0.5%. This is important.

  24. The issue is never to be found in absolute numbers,but in % expressions. 2.8% is low ,but could be lower be,or higher.Were 2.8% to become 4.3% say with a year then you have a 53% move. Such a % move when when either debt is high,and or real increases in net disosable income are low is an absolute nuke to consumption.
    It’s the % and the probabilities of a move that count. Back in the 70’s etc was actually brilliant. The rates were very high so the % move up from such alevel was very low and the probability of such a move was low because rates were already well above the long run average.
    Here we are below the long run average and the % move is correspondingly higher.Only amatching of rising rates with real increases in labour rates will prevent real economic problems.

  25. You are being human and ascribing a pattern where there isn’t one. Moving from silver to gold standard, war, going off gold standard, bank deregulations, .. all of this has much more effect than some number of years.

  26. True! Not to mention that the data series isn’t consistent. There was a certain amount of splicing necessary to put together a series that long.

  27. Dan, most corporate bonds these days have “make-whole” calls. This means that they are callable at a certain spread rather than a price. The call spread is normally much tighter than the current market spread, which means it would be very onerous for the issuer to call the bonds. Thus, the call is rarely triggered. If it the bonds ARE called, it would be a bonus for the bond holder because they would be taken out at a tighter spread (higher price) than the current market. It is kind of like a takeover in stock land.

  28. I would plan for rates going close to zero and staying there for a long time, based on the size of the outstanding debt (the Fed won’t issue bonds at 5 percent if it needs to borrow $2 trillion per year *and* rollover another $10 trillion in the next 5 years) and also the Fed moving closer and closer to outright monetization.
    At the same time, cost of living inflation will continue to rise.
    Very difficult to hold T-bonds in that kind of environment.