No, the “bond guys” weren’t right
I just heard several people on CNBC indicate something to the effect that “once again the bond guys were right, and the stock market was wrong.” This was in reference to the fact that the Fed’s disappointing policy announcement caused stocks to fall while long term bond prices rose.
Of course this is nonsense; both markets had the same set of expectations. Stocks fell because money is tighter than expected. Long term bond prices rose because money is tighter than expected. Operation Twist being extended was already priced in. Both markets thought something more was quite possible. When it didn’t happen both markets adjusted in exactly the way you’d expect.
Update: JMann points out that I probably misread CNBC. They probably didn’t mean the bond market was right, but rather those who were long in bonds were right. Nevermind.
Tags:
20. June 2012 at 08:49
It’s like they don’t even know how efficient markets are supposed to work… where do they get these guys?
20. June 2012 at 08:52
The more I see these ridiculous Fed statements, the more Morgan seems to makes sense. There are so many contradictions and hypocrisies in all these statements that it really does seem that the Fed is just looking for excuses not to do anything.
And what reason could they have for doing this? Getting Obama out of office seems as good a reason as any.
20. June 2012 at 08:54
Hi Scott
When the RBA cut rates in Australia recently long term bond prices fell, the stock market was flat and the AUD rose. The markets predicted a 50bp cut but they only cut by 25. Can’t the immediate reactions be a bit conflicting sometimes?
20. June 2012 at 08:56
Well, that’s what you get for watching CNBC.
20. June 2012 at 08:59
It is infuriating to be in a world where people think the only measures of monetary stance are the rate or money supply. Trying to argue this point with my fellow traders is like banging my head into a brick wall.
20. June 2012 at 09:00
The “bond guys” were right, and so were the “equity guys.” Shockingly, the CNBC talking heads are painfully wrong. Unfortunately they’re joined by most economists and most of the investment strategists that speak on CNBC. Rick Perry was on yesterday with Maria Bartolomo and I could barely restrain my rage at his complete idiocy.
Maria: Operation Twist is ending, so if the Fed doesn’t replace it with something else it’s effectively tightening policy. Do you think the Fed should tighten now?
Perry: HERP DERP FED SHOULD STOP PRINTIN MONEY. IT’S WRONG! THE IDEA THAT THIS WILL CREATE WEALTH IS WRONG!!1
If I was feeling optimistic I would note how perceptive Maria’s question was, rather than the stupidity of Perry’s response.
20. June 2012 at 09:00
heh. my favorite so far is “Goldman Sachs recommends buying straddles on XLV (the healthcare ETF) to take advantage of volatility surrounding the imminent SCOTUS ruling.”
right. its not priced in.
20. June 2012 at 09:04
Scott: does your title say what you want it to say? Wouldn’t a better title be “No, the “stocks guys” weren’t wrong”?
20. June 2012 at 09:09
Cthorm, actually, I think everyone was laughing except you.
Randy, if you’re a trader, and you’re right and they’re wrong, doesn’t that mean you profit? Why are you complaining?
Nick, love the new redesign!
20. June 2012 at 09:13
ssumner:
Long term bond prices rose because money is tighter than expected.
I don’t know about you guys, but reading Sumner posts is a source of endless amusement.
First he bites my head off after I said interest rates fall when money is looser (because he was defending his “interest rates rise when NGDP rises” story), and now he says the exact opposite “interest rates rise when money tight” (because he is defending EMH).
Up is down, black is white, and this blog is hilarious.
So if tighter money makes long term interest rates fall, then doesn’t that mean it is possible the loose money 2000-2007 made long term rates fall, specifically rates on mortgages, which then…dun dun duuun…brought about a housing boom?
20. June 2012 at 09:15
Actually, I’m the fool who is the source of amusement. I thought Sumner said long term bond rates, not prices.
I’m laughing at me right now.
20. June 2012 at 09:19
I’m more in the economist camp, but this seems to me to be an error in trader/economist translation more than anything else. I understand CNBC to meas that if you were long bonds and short stocks prior to the announcement, you made money.
1) Scott hears “bond guys” and translates it to “the price reached by buyers and sellers in the bond market prior to the announcement,” but I’m pretty sure CNBC means “people who thought that it was smart to buy rather than sell bonds over the past few days.”
2) I am in Scott’s camp on the EMH, so I think people who were long bonds were fortunate but not particularly right given what they knew prior to the announcement, but obviously a channel that specializes in information for non-passive investors is not going to be a big EMH believer.
20. June 2012 at 09:34
Tim, Yes, there are times when it’;s hard to read the immediate reaction.
Talldave, I just watched for 5 minutes, to get the Fed announcement.
Nick, I suppose that would have been better, but I meant to indicate that I disagreed with their claim that the bond market was more right than the stock market. It was illogical on so many levels I’m not sure either title would have been exactly right.
JMann. Yes, maybe I misunderstood them.
20. June 2012 at 09:39
JMann, I added a correction.
20. June 2012 at 09:39
ssumner:
JMann points out that I probably misread CNBC. They probably didn’t mean the bond market was right, but rather those who were long in bonds were right. Nevermind.
After my epic misreading of what you wrote, I don’t think anything you said would have been as bad.
20. June 2012 at 09:44
MF JUST ADMITTED HE WAS WRONG ABOUT SOMETHING!!!!
http://www.youtube.com/watch?v=4Fu8YIG8uyQ
20. June 2012 at 09:50
Saturos:
PRAISE JEBUZ!
I just became a means for you to tolerate every mistake you have ever made. Isn’t it awesome that we can feel better about ourselves in this way?
20. June 2012 at 10:17
Saturos, Sadly the markets I work in are not impacted are not directly impacted by this in the short term. Much more about risk mgmt then taking a position anyways.
20. June 2012 at 10:19
It seems that immediate stock market reaction to the Fed announcement was either extremely short lived, or there was some good news that quickly followed, because the market’s about where it was in the several hours prior.
I wouldn’t be surprised if markets weren’t too surprised by the announcement.
20. June 2012 at 10:24
Bernanke presser:
There you have it – stated clearly. He believes the total STOCK – of assets on the Fed balance sheet determines the LEVEL of accommodation. He is a bit mixed up. Expectations play absolutely no role in his approach.
20. June 2012 at 10:26
And now I see on CNBC a headline that the Fed lowered its economic growth forecast.
If the Fed:
a. believed it had the power to spur a full-recovery, and
b. was determined to bring about that recovery,
their forecast would be for a full-recovery.
This is like an episode of the 3 Stooges, except there are more stooges.
20. June 2012 at 10:26
He believes the Fed has many more tools – but – lets not use them yet.
What is in his mind? Does he have a mind?
20. June 2012 at 10:30
Why doesn’t anyone ask about NGDP targeting? Do none of these reporters read the blogs – or read Romer for example. This is like living in some horrid dream.
Can this go on for longer?
You bet it can – he says.
20. June 2012 at 10:34
Vast majority of reporters asking why the fed isn’t doing more… certainly a good sign.
20. June 2012 at 10:36
Before the Fed markets were mostly expecting about what they got-which is bearish.
The economic forecasters were mostly saying that we’ll just get more Operation Twist but that we need more.
The Fed as usual wanted err on the side of caution and so the sell off. Many market guys are saying now though that they read Bernanke’s commments as saying more is coming-that we will see QE3 soon.
20. June 2012 at 10:37
It is true that sometimes teh market will in start out thinking some new wrinkle of a plan out of Europe is great news until they notice that bond yields are rising in Spain and Italy, etc. and then fall.
20. June 2012 at 11:04
The “market” consists of many players with heterogeneous expectations. The market reaction on average is mostly right, but the whips are different players expressing their opinions.
A large number of Wall Streeters think that monetary stimulus is harmful to the economy, therefore it was inevitable there would be at least one short-lived rally on the lack of stimulus.
20. June 2012 at 11:08
Scott, Bernanke’s reaction to BoE’s funding for lending program was interesting. I felt he is supportive, and envied the fiscal support King will get from the treasury.
Bernanke is always talking about the risks and costs of the unconventinal tools, bad thing Obama does not want to pay for them.
20. June 2012 at 11:50
Long bonds rallied because… the Fed is buying *less* of them than expected??? The more they buy long bonds vs selling short bonds, the more long bonds will sell off?
I have a far simpler explanation, that’s actually consistent with basic rational asset pricing models, and the laws of supply and demand. Treasury bonds are negative beta. When the Fed unexpectedly removes them from investor portfolios, overall portfolio risk *increases*. I.e. bonds are a good hedge against real assets. Taking them away makes investors *less* able to take market risk. So yes, the announcement was “tighter” than expected. But not because they did too little. It’s because they did *more* than expected, as evidenced by the *rally* in the assets they are buying. Nobody wants operation twist. Somebody tell them to stop it, now!!!
20. June 2012 at 11:55
Steve:
The “market” consists of many players with heterogeneous expectations. The market reaction on average is mostly right, but the whips are different players expressing their opinions.
How can a domain of heterogeneous expectations be considered “right” or “wrong” in the singular?
This reification, if not anthropomorphism, of “the market”, as if it thinks, acts, and makes choices, should be abandoned.
A large number of Wall Streeters think that monetary stimulus is harmful to the economy, therefore it was inevitable there would be at least one short-lived rally on the lack of stimulus.
This comment is related to Goodhart’s Law:
“Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes.”
In other words, if NGDP targeting is adopted, then the historical information that NGDP theorists claim “confirms” their theory that NGDP explains macroeconomic phenomena like output and employment, will tend to collapse, and NGDP itself would cease carrying the information content that would qualify it to play the role.
It’s chasing a new reified concept that once controlled, makes it no longer able to do what it ostensibly used to do when it wasn’t targeted.
The same thing has occurred for every previous Fed target. The Fed targets bank panics, and bank reserves lose their information content. The Fed targets price levels, and price levels lose their information content. If the Fed targets spending, the same thing will tend to occur to spending.
Just like entrepreneurs NEED the information content in relative interest rates (which is distorted with controlled interest rates and thus the business cycle), so too do entrepreneurs need the information content in relative demands (which is distorted by inflation into distinct points in the economy, as a means to control aggregate spending).
NGDP targeting is not innocuous. It is not able to transcend individual action and enter some abstract world, after which entrepreneurs pick the floating dollars out of that abstract world, and into their business activity. No, inflation, since it enters the economy at distinct points, ALTERS relative demands, and thus prevents coordinated investment-consumption behavior among market actors, which leads to a building up of cash preference, which requires accelerated inflation in order to overcome in order to influence market actors into spending more that leads to 5% NGDP growth.
Hence, the (aggregate) money supply will accelerate as spending itself loses its information content and misleads entrepreneurs into systematic errors that can only be postponed and kept “liquid” by destroying the economic calculation attribute of money.
We have seen this in Australia 1990-2008. Acceleration in aggregate money supply growth cannot be sustained long term. The reason why the aggregate money supply growth accelerates is because the economic calculation nature of money is systematically destroyed, as what would have been a fall in aggregate spending that accompanies coordination, is overruled by inflation.
20. June 2012 at 12:04
JimP,
Here’s the clip that supports your assertion:
http://video.cnbc.com/gallery/?video=3000097557&play=1
That’s an incredible statement, especially in the face of the evidence.
20. June 2012 at 12:17
“Hi Scott
When the RBA cut rates in Australia recently long term bond prices fell, the stock market was flat and the AUD rose. The markets predicted a 50bp cut but they only cut by 25. Can’t the immediate reactions be a bit conflicting sometimes?”
I wouldn’t reason too much from central bank action when the country is outside of the zero lower bound. In “normal” times, a central bank has clear political power to essentially target NGDP. There are some digressions from NGDP targeting when there is supply-side inflation, but in general what’s considered high unemployment outside of the ZLB will force the central bank into easing sooner or later.
The big market movements based on future demand expectations only really happen in the ZLB. Or, in the case of the interwar period, when the gold standard similarly constrains central bank action. In that case, there is not clear communications for future demand and markets respond rapidly to some inkling of change in FOMC policy.
The only reason this divergence exists at the ZLB and not outside of it is persistent myths with the effects of central bank policy. Most macroeconomists still reason from interest rate changes. They only think in terms of the effects of lower interest rates. Milton Friedman successfully fought against this view for fighting high inflation, when he showed that double-digit interest rates could still be loose policy. The issue now is that, despite Milton Friedman’s (correct) advice to Japan, central bankers are still have a 1976 understanding of central bank policy when they’re in the ZLB.
20. June 2012 at 12:58
It is clear that the FOMC is doing everything it can to bring about Obama’s defeat without most people catching on about what it is doing. Token actions that do virtually nothing but give the impression that the Fed is trying to do something to carry out its maximum employment mandate is the way to achive this objective. Bernanke appears to be playing a solid part of this action.
The Keyenesians’ argument that we are in a liquidity trap plays right into the hands of the FOMC in helping them get away with this. After all, if the Fed has run out of bullets and expansionary monetary policy is like pushing on a string, one cannot really blame the Fed too much for not complying with its maximum employment mandate. And it diverts progressives from agressively attacking the Fed for failing to do its job, into pushing for fiscal stimulus policies that do not have a snowball’s chance in hell of getting through the current congress.
If Obama survives this cabal, replacing Bernanke as head of the f
Fed has to be his highest priority. On the other hand, if Romney wins, the Fed will switch to a more expansionary monetary policy.
20. June 2012 at 13:02
This is what front-running the Fed looks like:
http://i.imgur.com/3hV66.png
This is what Bernanke thought would happen:
“When our policy response lowers interest rates on govt bonds, it induces market participants to take more risk. Someone selling their govt bond to us may go out and buy a corp bond, thereby lowering spreads. A bank selling their govt bond to us may go out and make a loan…”
And we’re supposed to believe the 2003 Ben “I ♥ NGDP” Bernanke knew what he was talking about? The guy has been wrong about everything.
20. June 2012 at 13:20
“central bankers are still have a 1976 understanding of central bank policy when they’re in the ZLB.”
We are currently not is the ZLB (except with respect to the federal funds rate, but pegging the federal funds rate is only one, of a number of different ways, of conducting monetary policy).
To be at the ZLB, ALL interest rates that significanlty affect aggregate demand (and, in a dynamic context NGDP growth) have to reach a level (not neccessarily at zero) below which they cannot go. And that is currently not the case. Most obviously, as Scott has often pointed out, the rate on excess reserves can be reduced to zero. And that can be done by BOG edict. The presidents of the individual fed reserve banks do not have a vote on this. And by imposing a penalty on excess reserves this rate can be made negative. Also, mortgage interest rates are still a good distance above zero and can be brought down by agressive open market purchases. Pushing mortgage rates down to a level that virtually everyone considers unsustainable and then gradually letting them rise would have a powerful effect on AD. People who have been on the margin of buying or not buying real estate would rush to take advantage of the unusually low rates before they are gone.
20. June 2012 at 13:30
FEH,
What you suggest makes sense, but I just can’t accept the idea that the Fed changes course in the event of a Romney win. It was just last fall that Republican congressional leaders sent off this letter to Bernanke. http://blogs.wsj.com/economics/2011/09/20/full-text-republicans-letter-to-bernanke-questioning-more-fed-action/
It wouldn’t only be the Fed that changes course, but also the monetary stance of the Republican party at large. If THAT happens, then you truly have the Fed conspiring with Congress to sink a Democratic presidency. I don’t… I can’t believe that it is really playing out like that. Bernanke would then truly be known as the chairman who threw the independence of the FED away. I believe his own legacy would matter more to him than a single election outcome.
20. June 2012 at 13:38
“they read Bernanke’s commments as saying more is coming-that we will see QE3 soon.”
It is coming after it is too late to help Obama and in order to help Romney. We clearly have a political FOMC, but why are the other people Obama appointed to the BOG letting this happen?
20. June 2012 at 13:43
“Bernanke would then truly be known as the chairman who threw the independence of the FED away.”
Bernanke may well go down in monetary history as the Nicholas Biddle of the Fed.
20. June 2012 at 15:06
“To be at the ZLB, ALL interest rates that significanlty affect aggregate demand (and, in a dynamic context NGDP growth) have to reach a level (not neccessarily at zero) below which they cannot go.”
Well, now you’re just being an ass.
In all seriousness, your comment is correct and I honestly hadn’t thought about the ZLB that way. If the Fed set IOR at zero and then bought every last federally backed bond, then all risk-free rates would essentially be at zero because the only risk-free security left would be a Fed reserve account yielding zero interest.
That kind of shows my point that the zero lower bound in the Fed funds rate only matters for political reasons. Even if the (wrong) view that the Fed’s action only works through interest rates, why not reduce all risk-free interest rates to zero? Bernanke says more Fed action could be “destabilizing,” but why would long-term rates of zero be any more destabilizing than long-term rates of 1.4%?
Some might argue that long-term rates of zero or even negative w/ IOR penalty would encourage reckless lending. Bank lending, however, is not the only lever. If no investment opportunities exist, then those with money could just spend the money. Even if they do invest it wrongly, then the worst thing that will happen is that some money that would have otherwise sat on the sidelines would instead be spent on some dumb investment. That’s bad for the investor who shouldn’t have invested, but society as a whole is no worse for the money being spent vs. the money not being spent.
Anyway, all this to show that even at a naive microstructure level (expectations are the real lever in all this), the Fed is nowhere near out of ammunition and more Fed easing would not be destabilizing.
20. June 2012 at 15:11
Ultimately the market was flat whichi is appropriate as the market expectations were met but no exceeded. The median forecaster had expected Operatioin Twist.
Had Bernanke disappointed to the down side by not doing OT but just saying if things get any worse they’ll consider something the market would have tanked.
Had Bernanke overshot epxectations by doing soemthing beyond OT-QE3 maybe announced that they will keep interests low until unemployment hits 5.6% or somethting there would have been a rally.
20. June 2012 at 16:29
Scott – Five-year TIPS spreads have recovered to 188 bps from 165 bps in early June. They rallied today before and after the Fed statement and then sold off a bit (with the equity market) on the Bernanke press conference. I wish the Fed had done enough to push them to 300 bps, but at least they are going the right way in recent weeks. And the right way without ‘help’ from oil prices. Encouraging?
20. June 2012 at 16:58
Mike
That video link doesn’t run – at least for me.
20. June 2012 at 18:20
123, Yes, he seemed intrigued, But God knows why given he’s so worried about “risky” QE.
K, That went completely over my head. All I can say is the market reaction is no mystery. Lower NGDP growth expectations almost always make stocks fall and bonds rise, whatever the cause.
Matt, Good comment.
FEH, You said;
“If Obama survives this cabal”
Obama apponted 6 of the 7 members of this cabal, I mean Board of Governors.
Tommy, You have to be careful with small movements in the TIPS spreads, as they are strongly impacted by short term oil price changes. Once they show up in the CPI for a few months, they no longer depress the TIPS spread. But it doesn’t mean anything fundamental has changed.
Having said all that, I believe you are right in this particular case—as the other markets are also getting somewhat more optimistic.
20. June 2012 at 18:25
Just watched Mike’s clip from CNBC.
Bernanke said, “our view of the affect of the programs on the economy is that the total stock of outstanding securities in our portfolio is what determines the level of accommodation.”
Old fashioned monetarism is back, baby!
20. June 2012 at 19:35
“Obama apponted 6 of the 7 members of this cabal, I mean Board of Governors.”
As I stated in one of my above posts “but why are the other people Obama appointed to the BOG letting this happen?” One of the last 2 appointments was a Republican and had to be to make the Repulicans let the appointments go forward. But that leaves 4 more, including Janet Yellen. Why are they behaving like a bunch of wimps and not standing up to Bernanke? This really has me totally puzzled.
In any case, whether deliberate or not, the failure of the FOMC to comply with its mandate to achieve maximum employment is badly sabotaging Obama politically. One can disagree about whether this is good or bad, but that is what is happening. If Obama survives it will be in spite of the best efforts of the FOMC to do him in.
20. June 2012 at 19:44
“Obama apponted 6 of the 7 members of this cabal”
Actually he appointed 6 out of 12 members of this cabal. It is, after all, the FOMC that makes the monetary policy. And if you remove the recent Republican appointment and Bernanke, who was a great error, that leaves only 4 out of 12. Or if you take the President of the Chicago Fed into consideration, 4 out of 11.
But the failure of the other 4 Obama appointments to put more pressure for action on the FOMC is still a great puzzle to me.
20. June 2012 at 20:13
Scott,
“Lower NGDP growth expectations almost always make stocks fall and bonds rise, whatever the cause.”
If the Fed announces that rates are going to stay at zero for the next ten years, bonds are going to rise and so are stocks. So no.
“That went completely over my head.”
Then I definitely failed to be clear. The Fed announced that they are buying bonds from 6 to 30 years. Bonds from 6 to 30 years *rallied*. That means that the market was not expecting them to buy so many bonds. If the market had expected *more* buying of long bonds than they announced, then those bonds would have dropped on the announcement, right? So that establishes that the scale of the announcement was a surprise. *However*, stocks fell! They made a surprise purchase of bonds, and stocks *fell*. That is very damning of this policy.
What’s worse, there are excellent reasons based on rational asset pricing models why we’d expect operation twist to be contractionary.
20. June 2012 at 20:16
“JMann points out that I probably misread CNBC”
I doubt it. The nonsense knows no bounds.
21. June 2012 at 00:07
“If the Fed announces that rates are going to stay at zero for the next ten years, bonds are going to rise and so are stocks. So no.”
0% (or any rate) is consistent with both inflation and deflation. You need to know the intent of the policy, not just the path of interest rates.
That is what is frustrating about the Fed’s interest rate “predictions”. What do they really mean? Is the Fed saying, here’s some free money for interest rate arbitragers? Or is it some secret code for a future inflation target, because talking about inflation directly is politically impossible?
21. June 2012 at 00:30
Max:
0% (or any rate) is consistent with both inflation and deflation. You need to know the intent of the policy, not just the path of interest rates.
Bingo. The reason why 0% is consistent with any policy, is because the existence of interest rates is ultimately determined by individual time preference, not central planner dictates.
21. June 2012 at 03:41
K,
CNBC is dedicated to the idea that people can consistently beat the market using publicly available information, so most people here (including me) think they’re all wet, but once you watch them with that premise in mind, I think it’s usually possible to understand them. It’s like watching a channel devoted to acupuncture – it would be gibberish to a non-believer.
Scott,
Thanks – I feel famous!
21. June 2012 at 08:28
K,
Long-term bond prices are not based on nominal demand, but on future expected interest rates. If rates will likely stay around zero for the next six years, then it makes no sense to have six year bonds much above zero. If six year rates are, say, 3%, then markets can arbitrage the difference between constant 0% short-term rates and 3% six-year rates.
Whatever the Fed does as far as buying bonds, the market will not leave dollars on the table. If rates are more likely to increase in the future than suggested by the current lojg term rates, then the market will go short bonds until the rates adjust upward.
And when are future rates more likely to go up? When growth and inflation become more likely and the Fed then becomes more likely to raise rates in the future. Long-term bonds, in other words, *rallied* on news that the Fed did not seem committed to create future growth. This is not a reason for celebration.
Finally, stocks fell for the same reason bonds rose. They both have long-term stagnation priced in. Stocks fell because the expectation was QE3 and the Fed undershot that expectation. The fact that the Fed *undershooting* expected action caused stocks to fall is far from “damning.” In fact, precisely the opposite.
21. June 2012 at 10:41
Matt Waters,
“Long-term bond prices are not based on nominal demand, but on future expected interest rates.”
That is a very common, risk-free universe perspective. But it’s wrong. Bonds are risky by virtue of the fact that the path of the short rate is risky. If that risk is correlated with the market portfolio then the asset will have a risk premium. If holding the asset causes an increase to portfolio risk, the asset will trade at a discount to the expected payoff of a strategy which funds a purchase of the asset by rolling the funding at the risk free rate. If it decreases portfolio risk, it will trade at a premium to its expected payoff. Bonds are risky assets that currently provide a significant risk reduction to the market portfolio (which is why it makes sense to hold them even at slightly negative real returns).
Here is the point. If you reduce the weight of an asset in the market portfolio, you will *always* reduce the equilibrium return of that asset, i.e. you will increase the price. That is inescapable in any version of modern portfolio theory you will ever write down. I thought that part would be utterly uncontroversial.
The point I was trying to get at, which is more interesting, is that the impact on the price of the remaining assets depends on the correlation of the purchased asset with the market portfolio. If the correlation is negative, the market portfolio will drop in price, meaning that the remaining assets (stocks) will drop even more.
Dropping the yield by purchasing bonds and thereby reducing the weight of that asset in the market portfolio (changing the beta), has a *very* different, and *detrimental* effect, compared to dropping the yield by changing the outlook for the path of the risk free rate. If they had announced an even bigger program, stocks would have dropped *more*.
22. June 2012 at 09:10
K,
I feel pretty well-versed in CAPM and MPT, and I’m having a very hard time following your comment. I think you are forgetting that the foundation of those theories is the EMH. They say that current market prices exactly reflect future cash flows and risk premiums based on all publicly available information. If you don’t have inside information, then the best you can do is diversify away idiosyncratic risk in the market portion of your portfolio while putting enough cash in risk-free bonds to reduce your overall portfolio risk to an acceptable number.
What determines the risk-free rate in a world where the EMH applies? Well the EMH also says that no arbitrage opportunities exist where the profits exceed the transaction costs. You didn’t really make an argument about why the arbitrage argument is incorrect for long-term bonds.
In other words, it’s just not true that “if you reduce the weight of an asset in the market portfolio, you will *always* reduce the equilibrium return of that asset.” I’ve read over this sentence several times and still can’t really understand it. I do know it runs contrary to the EMH, which says future cash flows, risk-free rates and risk premiums determine the assets price. The CAPM assumes that however an investor changes their portfolio, it does not actually change market prices. Only news affecting, again, future cash flows, risk-free rates and risk premiums does that.
In this case, the news was that the Fed was going to take less action than expected. Due to the lack of arbitrage opportunities under the EMH, a decrease in long-term bonds meant a decrease in expected short-term rates. This pushed down the risk-free rate, which should have *increased* stock prices since cash flows were discounted at a lower rate. Why did stocks go down then? Because the Fed’s action signaled more economic stagnation in the future and, with economic stagnation, expected future cash flows went down more than the lower discount rate increased the PV of those cash flows.
22. June 2012 at 12:58
Matt,
“it’s just not true that “if you reduce the weight of an asset in the market portfolio, you will *always* reduce the equilibrium return of that asset.””
Matt! Take the basic CAPM with one systemic risk factor. Exogenous to the model are: The volatilities and weights of each asset in the market portfolio, and the correlation matrix that defines the joint normal asset return distribution, the risk free rate, and an investor wealth utility function. The purpose of the utility function is nothing more than to map total portfolio variance to risk premium. Given these parameters, the whole point of CAPM, then, is to compute the expected return
r_i = r + beta_i*(E[r_m]-r)
of each asset, i, given the risk free rate r, the expected market return E[r_m], and the asset beta beta_i. But *beta_i* depends not just on sigma_i and the correlation matrix, but critically on the weight of the asset in the market portfolio. Total variance of each asset clearly depends on quantity, right? Bernanke is changing the weights!
Consider a two asset model where the pairwise correlation is very negative (stocks and bonds). Let’s say total variance (w_i*sigma_i^2) of each asset is about the same. By symmetry, both assets will have equal and positive beta to the market portfolio. Now start removing the bonds from the portfolio. First of all, the beta of bonds will drop when the total bond variance drops. *That* will cause the bond yield to drop. And if you remove enough, the portfolio variance will soon be dominated by the stock asset because the bond weight is declining. Once the portfolio variance is stock driven, the bond correlation with the whole portfolio will be negative. At that point, if you remove bonds, portfolio variance will *increase*, which will cause the portfolio to drop in price (for the given risk aversion). Which is bad.
22. June 2012 at 13:54
Also, there’s an extensive literature on risk premia in the treasury bond term structure. In fact, the entire term structure literature assumes a term premium. You cannot hedge a term bond in the risk-free asset (the “bank account”).
22. June 2012 at 15:18
K, You said;
“If the Fed announces that rates are going to stay at zero for the next ten years, bonds are going to rise and so are stocks. So no.”
That’s not obvious to me—would you claim that’s also true in Zimbabwe? How about Argentina? How about Greece when they had the drachma?
You said:
“Then I definitely failed to be clear. The Fed announced that they are buying bonds from 6 to 30 years. Bonds from 6 to 30 years *rallied*. That means that the market was not expecting them to buy so many bonds.”
Wrong, Long bonds reflect future expected short term rates, which reflect future expected economic growth. Long bonds rally for the same reason stocks fell; the policy announcement was tighter than expected, and hence NGDP growth forecasts fell.
Max, I agree.
22. June 2012 at 19:58
Scott,
“That’s not obvious to me””would you claim that’s also true in Zimbabwe?”
No. I meant “*credibly* announces that rates are going to stay at zero for the next ten years”. Then it’s true.
“Wrong, Long bonds reflect future expected short term rates, which reflect future expected economic growth.”
If only it were that simple, Scott. As I told Matt, you can’t just ignore risk premium. What you are saying is that long bonds return the expected risk free rate. As if people are totally indifferent to investing in extremely volatile 30-year bonds and rolling their cash in t-bills for 30 years. People are not indifferent to risk. If they were, equities would also yield the risk free rate on average, and not 7% more, or whatever the equity risk premium is. The excess return of an asset increases with both the asset volatility and the correlation of the asset with the market. *Of course* bond yields are a function of the future rate dynamic (and I can write down that function if you want) but the bond yield is not *equal* to the expected average short rate. Rational asset pricing is quite a bit more complicated than that, and risk premium driven by volatility and market correlation are the critical determinants, as I discussed in my reply to Matt. You can’t just glibly dismiss 60 years of portfolio theory.
“Long bonds rally for the same reason stocks fell; the policy announcement was tighter than expected, and hence NGDP growth forecasts fell.”
I don’t actually disagree with that. But the policy that was too tight was *too much* operation twist, not too little. Not having worked through the basic theory is resulting in a *major* policy error. The least people need to do is actually understand the logical implications of MPT.
22. June 2012 at 21:14
“No. I meant “*credibly* announces that rates are going to stay at zero for the next ten years”. Then it’s true.”
In order to keep FF at 0%, they would have to take an ultra hawkish inflation stance, like the Bank of Japan. It’s only by constantly threatening to raise rates at the slightest hint of inflation that you can avoid having to actually do it.
If you say, well, the Fed should promise not to even threaten to raise rates – then we are in fantasy land because the Fed would totally lose control.
“As if people are totally indifferent to investing in extremely volatile 30-year bonds and rolling their cash in t-bills for 30 years.”
Not to sidetrack the discussion, but CAPM assumes that everyone is a “one period” investor. In reality, people invest over various time periods; a retiree for example may find a bond ladder to be less risky than cash.
23. June 2012 at 06:18
Matt Waters,
K’s argument about the bizarro portfolio rebalancing effect from QE and Twist (most people talk about it like it works the other way) via a CAPM is not unreasonable. For a fuller exposition of the idea and a somewhat thin attempt to show its application empirically see this paper:
http://www.federalreserve.gov/pubs/feds/2004/200457/200457pap.pdf
While plausible, I think both K and Scott are terribly overconfident in their respective, divergent explanations of the reaction to the Twist announcement. Portfolio effects (whether’s K’s or Bernanke’s) and Signaling (Scott and Woodford) have to interact in our jointly determined world. Plus, we could imagine very strange irrelevance effects countering a rebalance case that have nothing to do with assuming risk out of the utility function — maybe a reduction in government’s countercyclical liabilities (if you think long bonds are more countercyclical than short bonds) immediately reduces the forward beta on all government liabilities rendering no net effect on private portfolios. The bulk of the empirical evidence is all over the place (trying the above empirical study in other samples, or even extending the Japan sample slightly, doesn’t show anything convincing at all).
What happened in March of 09 when the QE announcement led to pretty significant increases in stock and bond prices? While both a signaling proponent and a CAPM rebalancing proponent could find reasonable ways to explain this reaction, the traditional rebalancing effect actually has to work a lot less to do so. But I don’t believe that one either. Frictions and expectations have to interact, making all but enormous market moves extremely difficult to untangle.
23. June 2012 at 08:00
K, I don’t agree that too much twist explains the tight announcement. I think it was failure to do other techniques.
Max, I agree.
24. June 2012 at 04:49
dlr,
It’s not *that* bizarro. I thought of it because that’s how I feel about my own PA. It was a gut reaction to the first twist. But I didn’t know about that paper. Thanks!
“the traditional rebalancing effect actually has to work a lot less to do so.”
Which one? The one that ignores covariance? I’m not *that* confident that I’m right. But the signature of the movement in bond prices did fit the exact announcement of which bonds they were buying and selling, so it doesn’t *look* like a response to a generic tightening. I’m not much for conspiracy theories, but the fact that the fed wrote that paper 8 years ago is actually making me wonder if they really are trying to kill the economy ahead of the election, as Krugman claims.
24. June 2012 at 19:31
Not sure if anyone is still reading this thread, but I think one issue I’m having is that I myself don’t really believe in CAPM. I didn’t post it before, but I was going to say that the CAPM would say a stock like Yahoo! demands the same risk premium whether in 1999 or 2001. According to CAPM, no investor should dare to guess that yahoo might be overvalued in 1999. Instead, the investor should glean the Beta of the stock from past price data, demand a risk premium for yahoo’s systemic risk and diversify away its idiosyncratic risk.
In fact, an investor actually looking at future cash flows would have easily steered clear of the stock market in 1999, when long-term Treasuries paid a lot more cash back to the investors than the sky-high P/E’s. Because it wrongly assumes the EMH in all circumstances, the CAPM has a lot of issues in actual empirical data. For example, higher Beta’s actually suggest a lower long-term return, the exact opposite relationship suggested by CAPM. Under the framework of robotic utility-maximizing investors, this relationship makes little sense. But it makes perfect sense if one assumes that some stocks become overvalued in good times and thus have a >1 correlation with the general market. However, in the long run, it will have lower return despite presumably higher risk. (There is an alternative explanation that higher Beta’s mean higher leverage and thus lower returns, but that doesn’t really help the EMH’s case either. Shouldn’t a rational market adjust prices downward of company’s leveraging out of cash flow issues?).
So, anyway, this is how I have understood CAPM. That’s why I am honestly confused by K’s arguments on how the CAPM relates to long-term risk-free rates. If CAPM holds, then long-term bonds would have a negative Beta. They are countercyclical as ultimately Fed policy decides the price movement of bonds and the Fed will vary interest rates inversely to the performance of the economy and the stock market. For this reason, bonds have rallied in poor economic news and fell in good economic news as long as I have followed the markets.
All the CAPM business just confuses the issue, which is why I don’t like it. Cash flows are still king and the arbitrage mechanism is through real cash flows from Treasury interest payments instead of just price movements.
BTW, as evidenced by my statements on the EMH, I am a good bit less dogmatic on the whole expectations thing. Expectations can help the Fed policy, but they may also have to do the dirty work of increasing demand through expansionary policy if the market irrationally does not believe them. For example, the market did not always front-run the Fed Fund rate announcements and the Fed would have to actually go down the bid or ask list until rates went to where they wanted. The Fed may still have to do that today, no matter how loudly the Fed announces they will expand policy until they move the NGDP meter.
25. June 2012 at 11:09
Matt Waters,
I’m still reading.
“But it makes perfect sense if one assumes that some stocks become overvalued in good times and thus have a >1 correlation with the general market. However, in the long run, it will have lower return despite presumably higher risk.”
There’s no pick in having a debate over irrational investor behaviour, if we haven’t first settled on the rational equilibrium. *That* is what I was trying to get at, and you were claiming that my argument was “wrong” because I didn’t understand the CAPM. Now I’m wrong *because* the CAPM is wrong?? Am I misrepresenting the CAPM, or not?
Please note that I’ve never argued my case based on a strict interpretation of the CAPM, but on general principles of MPT. My point is that bonds are a risk reducing asset *especially* under the the extreme circumstance (debt-deflation) where the normality assumption of the CAPM fails. I find the idea of a “protective” asset very intuitive.
“You didn’t really make an argument about why the arbitrage argument is incorrect for long-term bonds.”
“the arbitrage mechanism is through real cash flows from Treasury interest payments instead of just price movements.”
As I said, you cannot hedge the cash flows of a fixed rate bond in the risk-free asset. Please explain how to construct the arbitrage trade of which you speak.