This is not a thread about whether bonds are a buy or a sell. It is about what constitutes a bubble, and whether or not US bonds currently qualify under the accepted criteria for judging something a "bubble", namely stratospheric overvaluation that can't be justified by any plausible fundamentals-based valuation method.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Flaws in Warren Buffett's reasoning re: Treasuries "bubble"
- Thread starter Cutten
- Start date
This is not a thread about whether bonds are a buy or a sell. It is about what constitutes a bubble, and whether or not US bonds currently qualify under the accepted criteria for judging something a "bubble", namely stratospheric overvaluation that can't be justified by any plausible fundamentals-based valuation method.
Quote from harkm:
No argument will be convincing to you because you have already made up your mind and are completely fixated on the term "bubble". I personally believe that the 26 year bull market in bonds is over and this last rally was the final blowoff. Rates will go higher from here unless they are manipulated, which is certainly a possibility.
Any logical argument will convince me, as long as it is based on facts and reasoning. Feel free to offer one. Note that as a sceptic I *never* completely make up my mind about non-tautological statements (i.e. anything other than 1+1 = 2, 6 foot is taller than 5 foot etc). I would certainly never have a fixed opinion on any financial market.
Also note that my main point in this thread is totally independent of my personal position in bonds, or my opinion of their current valuation. I could think bonds are overpriced and yet still find Buffett's comments to be absurd and illogical.
Quote from Acumen:
It is only being called a bubble because anyone can tell its not going to last.
The target FF rate cant stay under 0.25% forever, when it goes up, the price appreciation T's have experienced will turn into a depreciation and the "bubble" will pop.
Its not rocket science.
TBT all the way.
Again - what position would you have had on in the 1930s, or in 1990s Japan? Short bonds, for the same reasoning? Most people then said it was obvious it couldn't last. You would have lost money year after year, just like they did.
Japan's equivalent of FF stayed below 1% for a decade and counting.
"Can't" is an expensive phrase in the markets.
Quote from scriabinop23:
why not. How long have Japan's overnight rates been at 1.0% or under ?
http://www.stat-search.boj.or.jp/ssi/mtshtml/d_en.html
That's an aweful long time to wait to be correct. Not to mention TBT can compound against you even if underlying goes flat, adding a virtual theta holding it. (yes I own some, and have experienced this, but I hold it as a hedge against a much larger variable debt liability.. so its worth it to me)
Yes, the cost of carry is painful for shorts if the correction takes too long. You could be right and still lose money.
Each year the bond market does not collapse, you lose 3%, compounded. If bond prices simply stay flat for the next 10 years, a short 10 year Treasury position will have lost 34% unleveraged by the time the bond matures. A long 10 year will have made exactly the same positive return. That is a difference of 68% of current capital - and that is unleveraged.
Ok, now you are just argueing semantics. Did you consider oil prices a "bubble" last fall? Most poeple didn't but oil dropped farther and faster than the Nasdaq crash. Who cares what this specific thread is about technically. It is now just a dicussion about bonds and their future direction. We will only know if long treasuries are in a "bubble"(again, who cares) in a year or two. We should bring this thread back up then and review.
Quote from harkm:
Ok, now you are just argueing semantics. Did you consider oil prices a "bubble" last fall? Most poeple didn't but oil dropped farther and faster than the Nasdaq crash. Who cares what this specific thread is about technically. It is now just a dicussion about bonds and their future direction. We will only know if long treasuries are in a "bubble"(again, who cares) in a year or two. We should bring this thread back up then and review.
http://scriabinop23.blogspot.com/2008/06/historic-analysis-of-crude-against-30.html
Funny isn't it? Look at the rest of the archive. I was not perpetually calling an oil top.
Quote from Cutten:
Many, many of the acknowledged prior investment bubbles, if not all of them, would easily qualify under my definition. What are your grounds then, for saying I am defining out the existence of bubbles? My post even mention at least two that have happened in the last decade.
I defined "bubble" in a way that I felt was consistent with almost all past examples now commonly accepted as bubbles, and that does not include what are not considered bubbles, that nevertheless had adverse price movement afterwards. If you think there are past bubbles that would not have met my conditions, perhaps you could list some? Or if you think some things would have fit my definition of bubbles but were not in fact real bubbles, again, perhaps you could give some examples?
I am interested in how you would define "bubble" - if my definition is wrong or misleading, please explain *why* it is wrong, and offer a better definition.
Okey your definition:
A bubble is when an asset becomes insanely overvalued such that even the most optimistic assessment of the fundamentals gives us a a valuation way lower than the current price.
Assume 7% inflation in average the next 30 years (not impossible) do you think bonds will not be considered to be in a bubble today?
If this indeed happen bonds today are priced on extremely low inflation expectations, and are therfore insanely overvalued.
The likelihood of a deflation scenario today will not change this.
Did everybody know the internet growth story could not continue in 2000 and beyond and that is why nasdaq was a bubble?
my guess is a bubble can only be agreed upon being a bubble in retrospect.
First, I commend Cutten on a tour de force (or tour de Treasuries). I like and respect most of his reasoning process, but do have five differences in data and logic that convince me that Treasuries are in a bubble.
1. Non-negligible default risk: CDS of 5-yr UST are now trading at 100 bp, up by a factor of 60 from historical levels of only 1.6 bp in just 2 years. With a first order analysis, this CDS price suggests a higher than 1% chance of default during each of those 5 years (more than a 5% chance of a default some time in the next 5 years). But this number underestimates the true default risk due to the existence of correlated counterparty risks (i.e, the very high chance that the CDS seller defaults at or around the same time that the US govt defaults during a global financial meltdown). Thus the real expected default rate for UST is much much higher than 1% and is probably is higher than the 2-3% that these T-bills yield.
2. Irrational T-bill-TIPS Spreads: In a rational market, this spread is a measure of inflation. In an irrational bubble market, the spread means nothing. That this spread was negative (an impossibility because TIPS will never return less than par in a deflationary environment) a few months back just shows that the spread does not mean what it seems to mean.
3. Oversupply issues: Currently the US government plans to run a 1.75 trillion deficit this year, a 1.1 trillion deficit in 2010, and more near-trillion dollar deficits into the future (assuming tax revenues don't fall further and assuming no additional bailout/stimulus spending). Thus, in the next 4 to 5 years, the total supply of T-bills will double. Other government will also be borrowing money, flooding the markets with sovereign debt. These future supplies do matter to the topic at hand. A bubble is measured not against historical ranges but against future expectation of price and returns. If I can buy a 30-yr T-bill in 2011 for 20% off the current price (or even 6% off the current price), I'd be irrational to buy one now. (The second failure of the German 10-yr Bund auction confirms the waning appetite for sovereign debt.)
4. Insane overvaluation is relative: For tulip bulbs, "insane" was 50X the long-term future price. For internet stocks, "insane" was maybe 4-10X the long-term price. For U.S. housing, "insane" was only about 2X the long-term price. For treasuries, the "insane" price is relative to the extremely low level of volatility and lack of speculative capacity (even the main street retail brokers let Joe sixpack have 20:1 margin on most Treasuries). If something is held for it's presumed extreme safety, then even the possibility of a 10% bubble loss is "insane." Thus, Treasuries can be within "normal" ranges but also be "insanely" overvalued. I suspect that when Buffett calls the bubble large, he's referring to the $ magnitude not the % magnitude of the bubble -- a 10% correction of prices of U.S. govt debt instruments would induce over $1 trillion in losses.
5. Rational expectations vs. plausible hopes: One of the most insidious elements of any bubble is that people can and do construct "rational" explanations for why the valuation for the item is "reasonable." We saw this clearly during both the dot-com and housing bubbles in which respected people (central bankers among them) had good reasons for why internet companies really would justify their sky-high valuations and why housing was really worth such a high % of people's income levels. Plausibility of valuation is no proof of the absence of a bubble.
Thus, the test of a bubble isn't whether the price is reasonable in the eyes of the optimists, but whether: 1) the statistically expected, risk-adjusted returns are positive; AND 2) the price is attractive to a sufficient number of realists and pessimists to absorb the total supply. The relatively high risk of default, high chance of inflation and the certainty of ballooning supply suggest that Treasuries are, indeed, in a bubble.
1. Non-negligible default risk: CDS of 5-yr UST are now trading at 100 bp, up by a factor of 60 from historical levels of only 1.6 bp in just 2 years. With a first order analysis, this CDS price suggests a higher than 1% chance of default during each of those 5 years (more than a 5% chance of a default some time in the next 5 years). But this number underestimates the true default risk due to the existence of correlated counterparty risks (i.e, the very high chance that the CDS seller defaults at or around the same time that the US govt defaults during a global financial meltdown). Thus the real expected default rate for UST is much much higher than 1% and is probably is higher than the 2-3% that these T-bills yield.
2. Irrational T-bill-TIPS Spreads: In a rational market, this spread is a measure of inflation. In an irrational bubble market, the spread means nothing. That this spread was negative (an impossibility because TIPS will never return less than par in a deflationary environment) a few months back just shows that the spread does not mean what it seems to mean.
3. Oversupply issues: Currently the US government plans to run a 1.75 trillion deficit this year, a 1.1 trillion deficit in 2010, and more near-trillion dollar deficits into the future (assuming tax revenues don't fall further and assuming no additional bailout/stimulus spending). Thus, in the next 4 to 5 years, the total supply of T-bills will double. Other government will also be borrowing money, flooding the markets with sovereign debt. These future supplies do matter to the topic at hand. A bubble is measured not against historical ranges but against future expectation of price and returns. If I can buy a 30-yr T-bill in 2011 for 20% off the current price (or even 6% off the current price), I'd be irrational to buy one now. (The second failure of the German 10-yr Bund auction confirms the waning appetite for sovereign debt.)
4. Insane overvaluation is relative: For tulip bulbs, "insane" was 50X the long-term future price. For internet stocks, "insane" was maybe 4-10X the long-term price. For U.S. housing, "insane" was only about 2X the long-term price. For treasuries, the "insane" price is relative to the extremely low level of volatility and lack of speculative capacity (even the main street retail brokers let Joe sixpack have 20:1 margin on most Treasuries). If something is held for it's presumed extreme safety, then even the possibility of a 10% bubble loss is "insane." Thus, Treasuries can be within "normal" ranges but also be "insanely" overvalued. I suspect that when Buffett calls the bubble large, he's referring to the $ magnitude not the % magnitude of the bubble -- a 10% correction of prices of U.S. govt debt instruments would induce over $1 trillion in losses.
5. Rational expectations vs. plausible hopes: One of the most insidious elements of any bubble is that people can and do construct "rational" explanations for why the valuation for the item is "reasonable." We saw this clearly during both the dot-com and housing bubbles in which respected people (central bankers among them) had good reasons for why internet companies really would justify their sky-high valuations and why housing was really worth such a high % of people's income levels. Plausibility of valuation is no proof of the absence of a bubble.
Thus, the test of a bubble isn't whether the price is reasonable in the eyes of the optimists, but whether: 1) the statistically expected, risk-adjusted returns are positive; AND 2) the price is attractive to a sufficient number of realists and pessimists to absorb the total supply. The relatively high risk of default, high chance of inflation and the certainty of ballooning supply suggest that Treasuries are, indeed, in a bubble.
Quote from Cutten:
No one has yet said why it's a bubble. All the arguments against have just said it's poor value. Let's take each in turn.
It does matter if it's a bubble. Why? Because bubbles eventually burst, *even if the good fundamentals stay intact*. High price alone is enough, when sustained for years, to cause a huge bust. Technology has continued to boom in the 2000s, the promise of the internet was arguably *underestimated*, yet here the nasdaq languishes 75% off its highs from 2000. Therefore if its a bubble, you should exit and not invest, even if you think we have 10 years of depression.
Whereas if you are a bear because you expect future inflation, or massive issuance, a la Pabst, then you can only be bearish *if those fundies stay intact*. In other words, if this slump turns into a depression, and there is huge demand even in the face of rising supply and debt/GDP ratios, then you have a direct challenge to your thesis in the form of fundamentals getting more bullish for the bond market. You can be flexible, because your position is dependent on a view of the future. With a bubble, it is a sell (at least in value investment terms) *under all possible future scenarios*. That's why it's an important distinction.
Secondly, a lot of people have been saying off-hand that the market is expensive. Historically the yields are low, no doubt. But a low yield does not necessarily imply expensive, as I pointed out previously. Under deflation, or even 0% inflation, 3% yields are reasonable for a low-risk safe haven asset. Once you throw in an unstable banking system and insolvent FDIC, 3% could actually be *cheap*.
All the arguments for inflation, supply, government spending, currency debasement etc were also present in the 1930s and 1990s Japan, times of long-term collapsing yields and soaring bond mega-bulls. FDR took the dollar off the gold-standard, massively expanded the federal government, spent and borrowed significantly, issued lots of new supply - inflation was non-existent and bonds kept going higher. Japan in the the 90s and early 2000s saw a lot of the same things - huge supply increases, huge spending increases etc. Yet inflation was non-existent and yields went lower and lower to 0.5% on the 10 year.
Here we are facing a similar banking and real-estate crisis. Volcker, Soros and others have made the comparisons. Since the inflation drivers did *not* cause inflation in the 30s and 90s, why would they this time? Since credit collapse and a historic real estate and stock market bust caused zero inflation or deflation in the 30s and 90s, why wouldn't it do the same this time? And if it does, why wouldn't bonds rally just like they did those other two times?
That's the question every bond bear - let alone the "bubble" crowd - have to answer. So far I have not yet heard a convincing argument.
Here is why it is a bubble. You are analyzing bonds as if they were a closed system, in the sense that their price is one-way function of variables such as inflation. However, the price of bonds (or their yields) can affect the variables you used as input to price them. The current price of bonds can cause a rise in inflation, and with it decrease in current bond prices. We need only the trigger for this.
The reason that bonds could be in bubble is that they can lead to higher inflation. The root reason for this is that the lower cost of borrowing can increase prices of assets, and later a decrease in the prices of bonds because of their current low yields.
This has not happened yet, because in assets such stocks, the far distanced earnings are expected to be negative, and given the current lower yieds, the causes their discounted values to the present to be greater in absolute value, which, if one adds the negative sign, causes a decrease in asset prices.
Once a bottom in earnings expectations, farther in the future, is reached and when these earnings expectations turn positive, the low interest rates will have the opposite effect on the valuations of stocks/other assets.
Once this takes place, bonds will be used as a borrowing source to take positions on those assets. Once the new fundamentals of assets will become appearent, asset prices will rise with an increasing rate. This should then increase inflation expectations, and a rise in bond yields.
If you couple this with the migration of cash from bonds positions to stocks/other assets, then the selling of bonds will intensify, and the inflation expectation will rise further, and you will clearly see at that time ( in the future) that bonds of today are in a bubble.
The flaw is then is in ignoring feedback loops which will kill bond's current bubble, and initiate a bull market in stocks, which would ultimately become a bubble itself.
Using current inflation to analyze current bond prices does not explain anything--- it is a tautology.
I do not know the reasoning of Buffet, but I think his conclusion is correct.
Im definitely not an expert in economics or bubbles so perhaps we can look at the problem visually?
If we were to overlay chart patterns of bubbles say
Nasdaq
Nikkei
Crude
Housing
with a chart of treasuries......
Could we then compare the similarities and differences and decide if treasuries are in a bubble or not?
If we were to overlay chart patterns of bubbles say
Nasdaq
Nikkei
Crude
Housing
with a chart of treasuries......
Could we then compare the similarities and differences and decide if treasuries are in a bubble or not?