I've always been annoyed by the term "risk-free bonds rate", meaning the return on US Treasury bills.  Just because US bonds have not defaulted within their trading experience, people assume this is impossible?  A list of major governments in 1900 would probably put the Ottoman Empire or Austria-Hungary well ahead of the relatively young United States.  Citing the good track record of the US alone, and not all governments of equal apparent stability at the start of the same time period, is purest survivorship bias.

The United States is a democracy; if enough people vote for representatives who decide not to pay off the bonds, they won't get paid.  Do you want to look at recent history, let alone ancient history, and tell me this is impossible?  The Internet could enable coordinated populist voting that would sweep new candidates into office, in defiance of prevous political machines.  Then the US economy melts under the burden of consumer debt, which causes China to stop buying US bonds and dump its dollar reserves.  Then Al Qaeda finally smuggles a nuke into Washington, D.C.  Then the next global pandemic hits.  And these are just "good stories" - the probability of the US defaulting on its bonds for any reason, is necessarily higher than the probability of it happening for the particular reasons I've just described.  I'm not saying these are high probabilities, but they are probabilities.  Treasury bills are nowhere near "risk free". 

I may be prejudiced here, because I anticipate particular Black Swans (AI, nanotech, biotech) that I see as having a high chance of striking over the lifetime of a 30-year Treasury bond.  But even if you don't share those particular assumptions, do you expect the United States to still be around in 300 years?  If not, do you know exactly when it will go bust?  Then why isn't the risk of losing your capital on a 30-year Treasury bond at least, say, 10%?

Nassim Nicholas Taleb's latest, The Black Swan, is about the impact of unknown unknowns - sudden blowups, processes that seem to behave normally for long periods and then melt down, variables in which most of the movement may occur on a tiny fraction of the moves.  Taleb inveighs against the dangers of induction, the ludic fallacy, hindsight, survivorship bias.  And then on page 205, Taleb suggests:

Instead of putting your money in "medium risk" investments (how do you know it is medium risk? by listening to tenure-seeking "experts"?), you need to put a portion, say 85 to 90 percent, in extremely safe instruments, like Treasury bills - as safe a class of instruments as you can manage to find on this planet.  The remaining 10 to 15 percent you put in extremely speculative bets, as leveraged as possible (like options), preferably venture capital-style portfolios. That way you do not depend on errors of risk management; no Black Swan can hurt you at all, beyond your "floor", the nest egg that you have in maximally safe instruments.

Does Taleb know something I don't, or has he forgotten to apply his own principles in the heat of the moment?  (That's a serious question, by the way, if Taleb happens to be reading this.  I'm not an experienced trader, and Taleb undoubtedly knows more than I do about how to use Black Swan thinking in trading.  But we all know how hard it is to remember to apply our finely honed skepticism in the face of handy popular phrases like "risk-free bonds rate".)  Regardless, I think that if you advise your readers to invest 90% of their money in "extremely safe" instruments, you should certainly also warn that it had better not all go into the same instrument - no, not even Treasury bills or gold bullion.  There is always risk management, and you are always exposed to error.  The safest instruments you can find on this planet aren't very safe.

New to LessWrong?

New Comment
40 comments, sorted by Click to highlight new comments since: Today at 10:36 PM

Most people who have thought carefully about the risk-free interest rate realize that any real-world security provides merely an approximation to that ideal. The fact that people rarely describe t-bond rates using verbose but more accurate phrases such as "the nearest we can come to measuring the risk-free interest rate" doesn't tell you much about how many fail to see that it's more accurate. I haven't read Black Swan (but have read a prior book of Taleb's). I doubt typical investors ought to follow the advice you've quoted, but it seems plausible that some investors ought to. His description of treasury bills as extremely safe seems accurate enough for practical purposes. It only requires that investors be able to anticipate a U.S. government default / hyperinflation something like 90 days in advance (i.e. it's a good deal more reasonable than describing 30-year bonds as extremely safe). Good investing is mostly about avoiding big mistakes, not about perfectly avoiding all errors, and Taleb's advice would reduce one's risk by a big factor compared to most competing advice.

The problem with anticipating a default 30 days in advance is if other investors anticipate the same - it will be too late to unload.

I don't think investors actually believe there is zero risk - the phrase "risk-free" is a holdover from the assumptions of theoretical asset pricing models, and now everyone who gets trained in economics or finance is just being sloppy.

It would take more than the USA ceasing to exist for bonds to become worthless - maybe the USA will peacefully disincorporate, but still unwind all its debts. It would take something that would destroy the productive capacity of the USA as a whole, or something that causes massive tax revolts - an asteroid strike, or an alien invasion maybe? If you're an optimist about the economic effects of AI/nanotech etc, then it seems that should make the chance of defaults less likely, not more.

Rather than some blowup like al Qaeda smuggling in nukes, a more likely scenario would be something much more mundane and stupid, e.g. a political gridlock between the President and the Congress, as happened briefly back in the 90s between Clinton and the Newt Gingrich-led House. Various government activities were shut down. It never got far enough to lead to not paying interest on the national debt, but it would not take all that much more of a conflict to do it, especially if we had pressure from abroad, with our humongous foreign debt. It has long been a ridiculous joke that Moody's downgraded the quality of Japanese bonds, when they sit on nearly $1 trillion in forex reserves, while US securities get top rating, while the US owes over $3 trillion abroad, a substantial chunk of that to the Japanese.

Upvoted for prescience.

Indeed!

[This comment is no longer endorsed by its author]Reply

Eliezer, your use of the term 'risk free bonds rate' is confusing. There is clearly no such thing a risk-free bond, so the notion of a risk-free bond rate doesn't seem to make sense. True, it's not impossible to lose money in short treasuries, but this has nothing to do with what's more commonly known as the 'risk-free rate' in modern portfoio theory.

The 'risk-free rate' is simply shorthand used to refer to a particular input to modern asset valuation models. It says nothing about a given security's total inflation-adjusted return, nor is it a conclusion drawn from a given government's past default history. The interest paid on short treasuries is considered risk-free in the limited context of asset valuation simply because a government can theoretically print as much money as it needs to pay the coupon.

That said, you're also correct to point out that in English the descriptor 'risk-free' suffers from imprecision. But this is just what happens when the shorthand names for inputs to mathematical models bleed over into the vernacular. That is, from a semantic standpoint 'risk-free' doesn't mean risk-free.

And I can certainly appreciate your poke at Taleb for not explicitly acknowledging potential black swan events affecting short treasuries. But as you know, he assumes you want positive returns on your money and is suggesting what he sees as the most optimal balance between risk and return in a 100% invested portfolio. Not that he's right or wrong about any of it.

I believe Egan-Jones once downgraded US debt to AA because of our balance of payments problems and our rather irresponsible spending habits. I suppose Taleb's tactic would make sense if one of the Swans you were betting on was a treasury default.

You'd only assume a 10% risk given a 30 year bond over a 300 year future if you assumed an even risk distribution over the 300 years.

Eliezer, Your criticism seems a bit superficial (in that I think almost everyone already realizes that T-bills aren't inherently "risk-free"), it's just that the U.S. federal government is perceived to have the best credit-rating in the world, in terms of likelihood to pay back money loaned to it.

Still I think it can be useful to be reminded of obvious things.

Eliezer, if you anticipate a default more than 90 days in advance, it doesn't matter that other investors do also. You hold the Treasury bills to maturity and they are paid off before the default.

Hopefully Anonymous,

That is according to US based bond raters. A lot of people think there is politics in that. Do you really think the US government is more reliable about paying back its bonds than the Japanese one? Really?

Barkley Rosser,

I haven't studied the issue, but I suspect that the "objective" determination is that the US Govt. is less likely to default on bonds than the Japanese govt. Really. Japan had regime change as recently as the 1940s. And as a country, it probably faces greater threats than does the United States, particularly from China and North Korea. However, Tbills and Japanese govt. bonds are most likely in an elite class, in terms being seen as relatively low risk places to invest money.

Anyway, risk of default aside, what of interest-rate risk? Isn't Taleb's advice faulty because of that?

Taleeb's advice is extremely faulty, at least for long run investing. The black swan event that makes a more fully diversified debt/equity portfolio fail disastrously in the long term (20+ years) has a fairly high likelihood of taking out a short-term US treasury portfolio as well.

More importantly, while it may fare slightly better on average in the event of a complete meltdown in riskier financial markets -- short of that, it will almost always do quite a bit worse. If the chances of that meltdown are, say, 10%, we have to weigh insurance value in that scenario (which as Eliezer mentions is hardly perfect) against being many times as well off the other 90% of the time.

Asking whether you've hedged against a US government default is a bit like asking "where's your end of the world insurance?" If the US government does default, chances are that it'll be because of a major disaster, meaning that most other investments will be worthless - along with a good chunk of the current rules of finance.

The only hedge I could think of would be a cache a guns, tined food, and petrol in a hidden (but nearby) location. Even if the US government default isn't due to the collapse of civilization, these are some things that would be guaranteed to maintain their value through a default.

Such caches and bunkers already exist (I suspect the whole 'zombie survival plan' fad helped popularize the concept) and some of them accept new applicants. It's a legitimate avenue of investment for anyone who's concerned about scenarios where modern infrastructure fails but the planet as a whole remains habitable.

If you buy a share of such a cache, what makes you think you would actually receive what you've purchased when the time comes, given that there will probably not be a functioning government to enforce property rights?

Schelling points. Only so many can fit in a given bunker, so some people must be excluded; excluding all those who did not sign up in advance, and only those, is a very straightforward place to draw the line.

Also, the group already inside the bunker is depending on commitments to each other. If the decision to exclude people is less than unanimous, there could be an internal breakdown of trust at the worst possible time.

That said, being able to reach the bunker entrance under your own power, bringing some extra supplies and/or blackmail materials along as bargaining chips, having a slot reserved at more than one independent fortification, etc. would be sensible hedging strategies.

Schelling points. Only so many can fit in a given bunker, so some people must be excluded; excluding all those who did not sign up in advance, and only those, is a very straightforward place to draw the line.

Another obvious Schelling point is to exclude all but those with the most political, social or physical power (whichever of those seems to be the dominant factor at the time). I don't particularly like that Schelling point but it is the one humans follow by default.

That's why I mentioned hedging strategies.

Even if having signed up in advance is not actually a sufficient condition for entry when the time comes, it can still be useful. Whatever coercion you manage to apply, the guard can rationalize by thinking of it as a matter of simply honoring your original reservation.

(I suspect the whole 'zombie survival plan' fad helped popularize the concept)

The Cold War seems like a far better cause for Doomsday preparedness. I wonder if, when the risk of nuclear war lowered, all the built up worry transferred over to "zombie survival," as that's something that I haven't quite figured out the psychological motive behind.

[-][anonymous]12y40

According to The Last Psychiatrist:

The zombie becomes the externalization of the ambivalence of mourning.

Yes, it's psycho-speak, but if you read the article he breaks it down a bit better, and even explains what the transition point between Cold War zombies and modern day zombies was (in his opinion).

I'm a fan of TLP, but I think my question is somewhat different. Zombie media and daydreaming about zombie survival strike me as very different things. If someone buys a gun or an axe because they think it'll be useful when zombies strike- what's going on there? It doesn't seem like disguised preparation for a race war, say, because there doesn't seem to be an actual element of fear, just a mimicking of fictional people who have their acts together. Maybe humor is relevant- if it's "fun" to plan what'll happen when the world ends, that seems different from sober-faced worriers building bunkers.

Hm.

The risks in government bonds aren't only those of default. Inflation (or exchange rates if you hold foreign bonds) can reduce the real value of bonds.

tc said "If you're an optimist about the economic effects of AI/nanotech etc, then it seems that should make the chance of defaults less likely, not more."

This is correct, but what are you holding the bonds for? Suppose during the next 30 years we experience a technological singularity. Your 30 year bonds will pay out as you expect, but the money will buy you relatively much less stuff; you might even be very poor by contemporary standards. (Incidentally what would the best investment/consumption strategy be if you thought a singularity was imminent? Consume only the barest minimum, and invest everything in stocks or other business opportunities? If many people behaved like this interest rates should go to zero, or even become negative.) This argument applies to any situation where economic growth is higher than expected. Your bonds payout, but the relative value of the cash is less than you expected, and you have a lower standard of living than you hoped - you can't eat at as good restaurants, you have to live in a less nice area; although you could partially compensate for this by investing more in other positional goods like playing music, writing poetry, studying mathematics etc.

What happens to the price of bonds if Aubrey de Grey and friends manage to cure ageing? Pension funds and life insurance companies are one of the biggest investors in bonds, so this is bound to affect bond prices. What happens to the price of land in such a scenario?

What about a collapse of the Euro, or a breakup of the EU - possibly as a result of Russian machinations. This seems more likely than US government default or a breakup of the US, and it would undoubtably affect the entire world economy.

Incidentally what would the best investment/consumption strategy be if you thought a singularity was imminent? Consume only the barest minimum, and invest everything in stocks or other business opportunities? If many people behaved like this interest rates should go to zero, or even become negative.

Hanson's simple models say that even under conservative assumptions, machine intelligence could increase annual world GDP growth rates to 25% from 3 or 4%. In that sort of steady state model, presumably the future growth would be priced into bonds or else investors would flee them for equities. On the other hand, if the market is insufficiently efficient/omniscient to foresee such an increase in growth rates, then there'd be a period where investors locked into low fixed-rate bonds will be screwed and missing out on the huge gains being reaped by equity investors.

I think markets are mostly efficient and I haven't heard of even long-term bonds being priced very highly, so I would guess that either no machine intelligence is in the offing or the markets are not being efficient about this. Since I have strong reasons to believe that the former is false, I choose the latter - the markets are inefficiently prizing the current low fixed-income offerings. Hence, buying equities would be a better long-term strategy.

Stuart Armstrong,

I think you are way overstating things. Did you read earlier comments? Are you aware how close we came to defaulting in the 1990s when Congress and the president gridlocked over a budget? We have a systemic setup in contrast with all those countries with parliamentary systems that could let it happen much more easily with nothing near some kind of horrendous catastrophe. And if we began to experience some kind of serious pressure from a foreign creditor, China anyone? on top of some internal deep conflict and economic problems (and US politics is not all polarized, now is it?), it could happen.

Regarding this business of the foreign creditors, there is an important reason why most Americans simply pay no attention to our mounting foreign indebtedness: up until now the that humongoug foreign net indebtedness has not resulted in a noticeable net deficit on the investment income part of the current account part of the US international balance of payments. However, as that foreign net indebtedness increases, this will inevitably change. When we start to see several hundred billions of dollars flowing on net overseas as interest on that debt, and the accompanying upward pressure on US interest rates start to negatively impact the US economy, there might even be an outright move to do what Argentina did to our bankers, and what New York state did to British bankers back in the nineteenth century regarding the debts for financing the Erie Canal construction: consciously default.

The bottom line on why we even fuss about the "risk-free rate" is that it is a necessary input to the standard formuli for pricing all kinds of options and derivatives via Black=Scholes and its variations. Something has to go in there, even if the wise really do understand that the underlying asset on which that rate is based is not really so risk free after all.

It's not much of a singularity if there aren't real economic gains. I see no reason that prices of tdoay's typical consumption goods should go up. In fact, most of them should drop precipitously. Pricing of positional goods will skyrocket and become unaffordable to those who didn't create enough value or capture enough rents during the transition. It's hard to say what will become of economics in a time of no scarcity, though. I think most "needs" will become essentially free and the economy will consist primarily of positional goods.

One could argue on this definition that we are already pretty close to such an outcome, at least in the rich world.

I think you are way overstating things. Did you read earlier comments? Are you aware how close we came to defaulting in the 1990s when Congress and the president gridlocked over a budget?

I know, and I read those. However, all the 20th century defaults I actually know about - Argentina, the mentioned examples of the Ottoman and Austro-Hungarian empire, many countries after a revolution - were accompanied by civil disorder. Does anyone have recent examples of a peaceful government default? (there are a few examples I've found in the 19th century, though they were generally partial or temporary defaults).

As for the US, I think it's easy to exaggerate, in hindsight, the closeness to a default. I very much doubt that a proper default - a sustained refusal to honour debts - would have come out of that.

I think you'll find Ottoman and Austro-Hungarian finances were pretty shaky. Turkey was known as the "sick man of Europe" for a reason. U.S. T-bills would have been almost as good as the Bank of England :-) I'd point out that the UK defaulted (for some values of defaulted) in 1931-2 and stopped payments in forex on some war debts.

Mexico defaulted (kindasorta) in the 90s without having any more violence than Mexican politics tends to involve. Russia defaulted on the GKO bonds in 1998, without any more violence than post-Soviet politics tends to involve.

Stuart,

Argentina recently defaulted and all but got away with it. Yes, there were demonstrations in the streets, but there are also on a regular basis in the US. What is your definition of "civil disorder" then? Elections were held, there was no revolution or coup.

In fact, the Argentine example might yet come to mind in the next decade if not-so friendly foreign creditors start to give the US a hard time. I think it was Keynes who argued that if you owe somebody a million dollars you have a problem, but if you owe him a trillion dollars, he has a problem (have changed the numbers to account for inflation). We could default and view it as "screw them" act.

Yes, there were demonstrations in the streets, but there are also on a regular basis in the US. What is your definition of "civil disorder" then? Elections were held, there was no revolution or coup.

My argentinian friends living there described scenes of intense chaos and mass lawlessness - but that might be just a selection bias.

Mexico defaulted (kindasorta) in the 90s without having any more violence than Mexican politics tends to involve. Russia defaulted on the GKO bonds in 1998, without any more violence than post-Soviet politics tends to involve.

I'm convinced! You can have defaults without disasters. I'll buy some gold and international stocks and bonds to go along with my guns and petrol :-)

We could default and view it as "screw them" act.

Actually, seeing the anemic response from creditors to the Argentina default, contrasted with the howls of "moral hazard" that accompagnied it, it's strange that countries don't default more often. Maybe a default is an admission of failure, so leaders don't want to do it - even if it would be best for their country. Or maybe international diplomacy is restraining them.

There is no such thing as risk free, but expected rate with the lowest possible variance can be found.

For simplicity, I like real global GDP growth as a measure of how one performs compared to all agregated invesment.

I'm with Stuart Armstrong: how do you hedge against a US government default? "End of the world" may be an exaggeration, but what investment is going to be in good shape if the US decides not to pay back the national debt? Everything takes a big hit. The risk is non-zero, but it practically unavoidable.

Or maybe there are investments completely insulated from that collapse. I can't claim global expertise.

Zubon, I think that was much more true in 1950 (when I think the us. was 50% of the world economy) than it is to day, post-Cold War, when the US might be under 20% of the world economy, and where appreciation of market economics and liberal government are widely appreciated. Thankfully world economic product has heavily diversified away from the United State. With regions like the EU and East Asia, I think even a particularly disastrous US collapse wouldn't take the rest of the world down with it.

I just wanted to point out the fact that foreign governments have defaulted on debt does not prove this point. With a US treasury security you are promised to be paid a certain amount of US dollars at a certain time. With the exception of certain inflation protected securities, no promises are made regarding inflation. The US government controls the ability to tax and PRINT dollars so I think the odds are next to 0 that the politicians would default on these debts when they could just print away their problem. The term risk-free asset refers to the default risk, of course there is still interest rate risk, inflation risk, reinvestment risk and others. In this context I feel that the term risk-free is for all intents and purposes accurate.

Credit Default Swaps on US Treasury bonds are currently trading at 16 basis points, up from 1.6 basis points when this was written.

As commenters at the above blog observe, "I'm trying to imagine a scenario where a Treasury bond defaults, but a CDS contract promising payments in the event of that default is still worth something" and "So if you went long the underling Treasuries in July, and went short by buying the CDS, you would have made gobs of money being both long and short the same asset. My mind is blown."

There turn out to be a lot of scenarios. You point out one way:

The United States is a democracy; if enough people vote for representatives who decide not to pay off the bonds, they won't get paid. Do you want to look at recent history, let alone ancient history, and tell me this is impossible?

This is exactly the scenario unfolding now with the Republican brinkmanship on the debt ceiling. No new debt = default. From Forbes "What a US Default, Downgrade Might Look Like" (emphasis added):

"If Congress fails to increase the debt limit, the government would have to stop, limit, or delay payments on a broad range of legal obligations, including Social Security and Medicare benefits, military salaries, and interest on the national debt, which is paid to big, market maker banks like J.P. Morgan Chase, Citibank, and others, not to mention the government of China, which is the largest holder of US government bonds overseas. Defaulting on those obligations, including coupon payments to bond holders, would cause severe hardship for the US economy. It would erode the historic legacy of the US as the safe harbor within the global financial system."

May I submit reading anything by William Bernstein to anyone interested in a math-heavy discussion of this? The book "the intelligent asset allocator" is the most math heavy of the set.

His best quote on this is "assume a very fat tail" i.e. assume statistically unlikely events are rather likely as a group.

I would also like to take a moment to praise Alexander Hamilton and his friends for setting the precedent that the US has never defaulted. We were awfully close after the revolutionary war.

Perhaps he says more in Black Swan - but I don't see the quoted text as describing anything as risk free.

Treasury bills - as safe a class of instruments as you can manage to find on this planet

Is pretty much true. Its the single safest instrument that bears interest. Now I would not follow the rest of his advice, for reasons you mention I think it would be safer to hold some other notes as well some precious metals (which aren't interest bearing but at least float with inflation).

The other factor here that you don't touch on is that a lot of people believe that a US Treasury default would be pretty much the end of our currency and economy; notes issued by Swiss banks couldn't be expected to hold up since everyone indebted to them has some exposure to US assets or to someone else who does. I don't know if this is true but I'm quite sure many people believe it.