Well, THAT's more like it. Yesterday's risky-asset lurch was pretty much what Macro Man expected from a more hawkish (or, more realistically, less-dovish) Fed statement. Whether it proves to be yet another buyable dip, or a state change as credit conditions deteriorate again, of course remains to be seen.
That the Fed chose to inject $41 billion in repos yesterday offers a hint about the likely solution to any concerns of renewed market seize-up, and we should all probably be prepared for similar operations over the next couple of months as the Fed (and ECB, and BOE) seeks to avoid a year-end liquidity crunch.
As a brief follow-up to yesterday's post on crack spreads, Macro Man constructed a crude (boom-boom!) two-factor model of US gasoline prices, with crude prices and crack speads (3:2:1 formation) as the inputs. Unsurprisingly, there is an excellent correlation, not least because the gas prices are on both sides of the equation. But it's nonethelsss a useful to for gauging the potential magnitude of gas price shifts on the basis of some simple assumptions. With crude at $95 and crack spreads at their 1 year average, for example, the model yields a gas-price forecast of $3.43. Hard to see anyon enjoying that kind of move but the Grinch!
Yesterday's risk asset weakness was partially sparked by renewed rumours of massive, massive losses at Citigroup, partially driven by a CIBC analyst's report. Now, while it is obvously difficult to ascertain how much of the rumour is true, it certainly seems as if the reputation of Wall Street executives has taken a beating recently.
Consider:
Chuck Prince, Citigroup: The man Jim Cramer loves to hate. He may still be dancing, but alas the venue would appear to be the shattered dreams and broken P/L's of his investors.
Jimmy Cayne, Bear Stears.
Macro Man will leave the more scurrilious rumours about Mr. Cayne to others, but when the WSJ is reporting that you spent much of the July crisis period playing cards and on the golf course, it does suggest rather less attention to detail than investors, clients, and employees might have a right to expect. Is it any wonder that Bear's stock price has gone into hibernation?
Stan O'Neal, Merrill Lynch.
Stan's performance is reminiscent of another Stan the Man, complete with extremely large losses. Evidently, it takes a special, virtually unique, talent to lose as much money as Merrill did under O'Neal's watch, as Stan was handomely rewarded for his departure. As an aside, did anyone else find the manner of O'Neal's departure as amusing as Macro Man? Stan "retired" from Merrill Lynch? Yeah, just like Julius Caesar "retired" from politics in 44 B.C.E.
Anyhow, today sees another sacrifice at the altar of statistical noise- US nonfarm payroll data. Regular readers will know by now that Macro Man is more interested in the revisions than he is in the headline number, which contains much more noise than signal.
Insofar as any aspect of the October data bears watching, the unemployment rate is probably the thing to look at, as it provides a measure of the rate of labour force utilization. Certainly yesterday's spending and income data did not suggest a consumer on the ropes: y/y wage and salary gains were reported up 7%, and real disposable income continues to grow nicely courtesy of muted gas prices (for now.)
Where will we be at market's close today? Your guess is as good as mine; however, the sharp increase in volatility may be serving as a warning to dial risk down until things get a bit clearer.
That the Fed chose to inject $41 billion in repos yesterday offers a hint about the likely solution to any concerns of renewed market seize-up, and we should all probably be prepared for similar operations over the next couple of months as the Fed (and ECB, and BOE) seeks to avoid a year-end liquidity crunch.
As a brief follow-up to yesterday's post on crack spreads, Macro Man constructed a crude (boom-boom!) two-factor model of US gasoline prices, with crude prices and crack speads (3:2:1 formation) as the inputs. Unsurprisingly, there is an excellent correlation, not least because the gas prices are on both sides of the equation. But it's nonethelsss a useful to for gauging the potential magnitude of gas price shifts on the basis of some simple assumptions. With crude at $95 and crack spreads at their 1 year average, for example, the model yields a gas-price forecast of $3.43. Hard to see anyon enjoying that kind of move but the Grinch!
Yesterday's risk asset weakness was partially sparked by renewed rumours of massive, massive losses at Citigroup, partially driven by a CIBC analyst's report. Now, while it is obvously difficult to ascertain how much of the rumour is true, it certainly seems as if the reputation of Wall Street executives has taken a beating recently.
Consider:
Chuck Prince, Citigroup: The man Jim Cramer loves to hate. He may still be dancing, but alas the venue would appear to be the shattered dreams and broken P/L's of his investors.
Jimmy Cayne, Bear Stears.
Macro Man will leave the more scurrilious rumours about Mr. Cayne to others, but when the WSJ is reporting that you spent much of the July crisis period playing cards and on the golf course, it does suggest rather less attention to detail than investors, clients, and employees might have a right to expect. Is it any wonder that Bear's stock price has gone into hibernation?
Stan O'Neal, Merrill Lynch.
Stan's performance is reminiscent of another Stan the Man, complete with extremely large losses. Evidently, it takes a special, virtually unique, talent to lose as much money as Merrill did under O'Neal's watch, as Stan was handomely rewarded for his departure. As an aside, did anyone else find the manner of O'Neal's departure as amusing as Macro Man? Stan "retired" from Merrill Lynch? Yeah, just like Julius Caesar "retired" from politics in 44 B.C.E.
Anyhow, today sees another sacrifice at the altar of statistical noise- US nonfarm payroll data. Regular readers will know by now that Macro Man is more interested in the revisions than he is in the headline number, which contains much more noise than signal.
Insofar as any aspect of the October data bears watching, the unemployment rate is probably the thing to look at, as it provides a measure of the rate of labour force utilization. Certainly yesterday's spending and income data did not suggest a consumer on the ropes: y/y wage and salary gains were reported up 7%, and real disposable income continues to grow nicely courtesy of muted gas prices (for now.)
Where will we be at market's close today? Your guess is as good as mine; however, the sharp increase in volatility may be serving as a warning to dial risk down until things get a bit clearer.
28 comments
Click here for commentsRegarding Fed injection yesterday. Actually there was a 1.5b reduction. 42.5 maturing vs. 41b added.
ReplyCayne always attends the three (ten day) North American Bridge Championships (spring, summer, and fall) each year (I have seen him at them many times).
ReplyObviously, it looks bad to find on vacation with the world falling apart, but I don't know what would have ended up different if he had spent twenty-four hours a day in his office during that time.
KG
O/T, but I couldn't help noticing your generous response in yesterday's post explaining "beta plus" strategy.
ReplyI've read here and here.
I have two genuine questions, I am not carping or criticizing.
1) Why did you choose to use spreads, rather than a ratio of the two yields? In other words, what I am missing by finding a 6% yield versus a 4% yield more significant than a 12% yield versus a 10% yield?
2) With your voluminous dataset, why did you set your bins up by equal percentile range (spread basis points)? In my Statistics 101 class, they encouraged me to set up the histogram with an equal number of samples in each bin, if I wanted to estimate probability. I know you're a competent quant, what am I missing?
p.s. Sorry if the html isn't working, I couldn't test it with the PREVIEW button on this blog.
KG, a couple of observations:
Replya) In any industry, a senior manager who chooses to pursue personal hobbies at the expense of work during times of his firm's distress is unlikely to look good
b) I doubt that many other BSC employees have the opportunity to take take that much time off come hell or high water
c) If his presence at BSC and leadership during the crisis (remember, he wasn't just not in the office, he was unreachable by email) wouldn;t have made a difference....why does the firm need him in its employ at all?
Index investor, good questions. I could have just as easily used yield ratios, which perhaps may have made a difference in the 70's.
Re: the historgrams, I find it useful to look at discrete nominal bin ranges to get an idea of what the sample distribution actually is. I personally find it useful to know that in region x, the strategy does/does not seem to work, but the sample size is not siginficant.
In general, my philosophy with this sort of thing is that it is very easy to over-engineer a sample and end up with a false sense of confidence over one's results. I prefer to look for 'sweet spots' with a large cluster of profitable data points to use as a signal.
It may not be to everyone's taste, but I am comfortable with it- and given it's me who's running the models, that's the most important thing!
Thanks so much for your response.
ReplyOne final (thats a lie!) question:
Say UST yield is 6% and S&P 500 earnings yield is 3% to 3.5%
How do you determine your allocation? Do you buy UST, or do you short S&P 500? Perhaps two positions (long UST, short S&P500), but then, how much weight to each position?
As an index investor, its easy for me, I just go long. I'm interested to know what a more dedicated investor like yourself does.
In the case of this relatively simple model, it would have no mechanistic position in either equities or bonds.
ReplyIt would then be incumbent upon my alpha-generating skills to derive profits from the market.
The point of the model is just to keep me in equities as long as it should be relatively "easy money."
At Goldman, Paulson, Thain and Thornton with alot of help from Mindich bought Spear, Leeds, Kellog for nearly $8 billion in 2000. Two years later it was worth maybe $1 billion. GS loses $7 billion on that trade and yet they are still "heroes" of the industry. O'Neal's bad CDO trade should be put in perspective. Goldman just had a much better franchise to hide the bad decsisons by senior management.
ReplyThe $41B in repos and the larger than usual repos since August are due to the Fed's redeeming $10B in Treasury securities ( in August ) from its permanent portfolio , which exist to adjust overall liquidity in line with the long term growth of the economy
ReplyWhile Goldman may have made mistakes, and I do wonder at why so many people think that they're infallible, that does nothing to detract from the fact that O'Neal pushed Merrill into subrpime turds well after the horse, as well as the returns, had bolted, anded ended up getting paid off with a package that would have made Dick Grasso blush with embarrassment.
ReplyClearly I misseed a trick with the Fed repo and concde that yesterday's exercise was not particularly sinister. However, given current credit conditions and a likely tigfhtening of such conditions into year end, I still look for a mini-1999 flood of liqudiity into the system.
MM, With B/E's already creeping wider, won't a '99 flood cause them to break right the hell out?
ReplyRJ
Perhaps, RJ. But then again, I think breakevens are chronically underpriced. The beatury of TIPs, of course, is that the market doesn't have to price them correctly for you to make money: the inflation accrual takes care of that very nicely indeed.
ReplyThe TOMO main event it appears was someone getting 20bn MBS away for 7 days.
ReplySome quick replies to your thoughts:
Replya) I agree, but (speaking only about bridge here), I would call it more than a hobby. He ranks as one of the top players in the U.S., and the tournaments have a significant social and professional aspect (meaning a lot of the top players make their living as traders - and do quite well at it).
b) He defines the culture. The bridge tournaments would take just three weeks a year (and maybe an Friday or Monday on either end for travel to those at locations on the other side of the country). Not impossible to attend those (I used to do it every year, myself :) ), but of course, I never needed to keep up an image as an important person in a public company.
c) He owns (or used to own) a billion dollars worth of Bear Stearns stock, and he seemed infallible for many years. Certainly the case for keeping him employed now seems far weaker than it did a year ago. :) As for the alleged inability to contact him, the bridge tournaments have fairly structured times - one session from 1300 to 1700 and a session from 2000 to 2400. He could have remained in nearly constant contact outside of those times.
I don't want to sound like a defender of what he did (taking the time off for bridge alone seems quite adequate, but piling golf on top of that..., well), but it does sound like the story could have presented things in a way to make them appear worse than the reality of the situation.
I don't know the real details of his connectedness during that time of course, but it just seems like another moment to filter what a reporter writes through my own experience.
Quite honestly, who cares?
ReplyIf you were a shareholder, do something about the principal-agent problem, or sell your shares. Otherwise, aren't the Walton heirs, the Evil Scaife and Paris Hilton better poster children for the anathema that is deeply undeserved wealth in the new gilded age?
Wcw, Paris Hilton, et al are undoubtedly leeches, and I frankly find public interest in them to be a tad puzzling. Then again, I am bemused by modern celebrity culture full stop, so I suppose it isn't a surprise.
ReplyHowever, Paris et al have not presided over the type of destruction of wealth that these gents, among others, have. In some way, they are culpable by either their actions or their inactions.
Does this sort of thing keep me up at night? No, of course not. But I don't mind having a little fun at someone else's expense...particularly when I have friends employed at some of these firms who are being materially impacted by some of the boneheaded (in)/actions of the gentlemen depicted here.
Also Paris is more transparent, will, c/o Blackstone, be richer, and gives (though this may be disputed in some boardrooms).. better head.
ReplyHmmm.....but has she f****d every single shareholder like Chuck Prince has?
ReplyIndex investor.
ReplyYou should search the CFA institute website on "Fed Model".
There is a quite comprehensive article on the market timing ability of this yeild differentiall(unfortunality i don't remeber the name of the author). It uses a very large data set, dating back to 1880ies. In short he shows that while this model was good from 60ies, earlier it was a complete disaster.
Such things shoud not be a basis for investment decitions. Good performance from 68 might have beed a "spurrios correlation".
There are valid objections to this tipy of market timing - equities are longer term (well beyond 50 years) and subordinate claims, for example.
It should be at best used as a additional decision-making tool.
Of course, there are legitimate criticisms to the criticism: the non-performing periods took place under different monetary, exchange rate, regulatory, and tax regimes, all of which impact the relative price of equites and bonds.
ReplyI'm really not sure how relevant the poor performance of a strategy is during years when insider trading was not illegal and there was no Federal Reserve System. Actually, I am pretty sure: it's not relevant.
I would agree though that a bond yield/earnings yield comparison should not be the sole basis of an investment strategy; however, the vitriol with which it is attacked is frankly a little puzzling.
Not metaphorically, and literally she would dispute that there could be any being f****ing comparison, and add that you gotta give a girl a chance
Replymm - re: the fed model
Replybad science is bad science, full stop.
the fed model deserves about the same cred as say, creationism, for both are based upon sky-hook arguments that rely upon "faith". Moreover, in both cases the the believer cannot prove beyond the flmsiest of cases, though neither can the skeptic cannot entirely disprove the assertions of its champions. N ot at least until the endo f time, or equity & bond markets or both. However, IMHO there are more than sufficient reasons to tread so cautiously that, as a predictive tool, it is rather useless.
For good predictive equity model building, go check out Steve Lecompte at CXOAdvisory blog. His REY and RTV models and the consensus between them capture the real essence of equity return drivers per Ibbotson and other robust long-term researchers. Arguably, he has shifted the problem to one's ability to predict inflation to gain an edge in model breakdown, but that is a much better point of departure than a bad model with a whole range conditional idiosyncracies.
I believe you;ll find that most who have contentions with the Fed model (Asness, Me) etc. are far better able to calmy and scientifically articulate their objections than the proto-typical supporters (Kudlow?) who are more likely to be vitriolic and ad-hominem in their zealous (remember Masada?!??) support of ostensibly bad science.
About all you can say is it's (quite literally) "good enough for government work", but as Congress, Mssrs Greenspan & Bernanke have shown, look where THAT'S gotten us!
C, the CXO rebuttal of the Fed model appears to rest on a straw man:
ReplyFed Model proponents argue that there is an equilibrium relationship between the earnings yield of a stock index and the 10-year government bond yield.
and and an ad hominem:
the author concludes that practitioners who use the Fed Model are simpletons.
AFAIK, very few people claim the Fed model tells you the 'equilibrium' level of stock prices.
What it does offer, however, is some sense of relative attractiveness of stocks and bonds.
And indeed, this study suggests that that is where the model has its greatest utility.
My usage of it is simply as a mechanism to get out of bad markets, an activity which this paper would appear to endorse.
The fact is that there are a number of factors- tax regimes, Federal fiscal situations, participants in financial markets- that can alter the relative attraction of stocks and bonds.
While it is true that corporate earnings are a different claim from Treasury coupon payments, it is equally the case that an equity principal is different from a Treasury principal.
Focusing on dividend yield is, it seems to me, an even more flawed construct than the Fed model. Yes, a substantial portion of historical equity returns have come from dividends, but that statement is descriptive rather than prescriptive.
There should ultimately be no difference between retained earnings and ditributed earnings on stock returns, assuming a firm has a steady marginal return on the deployment of retained earnings.
Suppose you have a 100 shares of a stock trading at $100, so $10,000 invested. It declares a 50% dividend, which reduces the stock price to $50. AS prudent invstors, we re-invest our dividend in another 100 shares of stock. WE now own 200 shares of $50 stock. The stock proceeds to double in value, returning to $100. WE now have $20,000 worth of stock, 200 shares at $100. Happy days!
How is that any different from the same stock paying no dividend? Assuming it is the same stock, it will still double in value, leaving us with 100 shares of a $200 stock. In other words, we'll still have $20,000.
As an aside, there's another way to describe the 50% dividend and re-invested described above. It's called a "stock split."
Macro Man , i did not mean to critisize the rationalle behind the yeild differentiall.
ReplyI really think that is a usefull thing to look at, just like pretty much everything on the market which has a large group of followers who influence the price action. And while crude this is good relative valuation metric.
Me thinks that the differetiall is a bit to simlistic. If we talk about the yeild curve, then the shape also matters, and simple data search shows that it is also statisticaly significant.
As for arguement on the data relevance - unfortuatelly we'll never know, imho.
Some reaction to the good U.S. employment number....These markets are really concerned about what is hiding on the balance sheets of these big investment houses.
ReplyAs you imply, religion can be helpful to comfort those in need of some explanation for events that they cannot otherwise explain. I accept this , although I personally subscribe to the Church of the Apothetic Agnostic. As such, veritable explanations for events are highly likely to reside elsewhere.
Reply"There should ultimately be no difference between retained earnings and ditributed earnings on stock returns, assuming a firm has a steady marginal return on the deployment of retained earnings."
ReplyAnd maybe it's exactly that assumption which isn't true.
You think CEO's don't have recency and overconfidence psychological biases?
Indeed, I'd say there is a strong selection function for getting there which selects FOR exactly those traits in top management multiplied by some very good fortune.
I.e. they think that their recent retained earnings was due to their wise decisions (i.e. investment versus dividend), and so re-double the effort in doing exactly that.
More bluntly, big pots of undifferentiated Other People's Money are very tempting to loot or waste in an entertainingly energetic fashion. San Francisco 1999 was a real party I hear.
Dr. C, if we leave the world of theory and enter the world of reality, we do of course need to account for the taxation of dividends, which, from a compounding eprsepctive, are a real bitch, and render stock splits economically preferable to large dividend payouts.
ReplyThanks to all of you for sharing your excellent comments about the Fed Model and other approaches emphasizing an examination of the relative attractiveness of stocks with respect to bonds.
Replydr chaos: I was particularly delighted with your "big pots of undifferentiated Other People's Money" comments.