now with 30% less insight than regular blogs

Friday, February 03, 2006  

OK, it's been 2 years and roughly 2 months. What has transpired?

The unemployment rate is down to 4.7%.

National Net Worth is at a record.

Productivity growth in the past 5 years has been the highest of all time.

The Iraqis have had three elections.

Bush's favorable rating must be like, 75%, right? No? What the fu

posted by Pete Harrigan | 12:57 PM

Saturday, December 06, 2003  

I have not been posting much recently, but there is no excuse for not linking to this piece by Steve Antler, in which he explains why the federal debt and deficit are not major problems.

There. I feel better now.

posted by Pete Harrigan | 4:55 PM

Friday, December 05, 2003  

The unemployment survey results are out and they point to an unbelievably strong economy. According to the BLS' House hold Survey, the total number of jobs grew by 589,000 in November, while the unemployment rate dropped to 5.9%.

Of course, that is not the whole story. To be complete, one has to mention the payroll numbers from the Establishment Survey. Those showed a gain of only 57,000 jobs, considerably smaller than expected. To report only one set of numbers is to tell only half of the story, usually the half people want to report.

News reports routinely mention only the numbers from the Establishment Survey, probably due to the incredible intellectual laziness of reporters. Why dig in to numbers they will never understand when it is so much easier to report what everyone else is?

Liberal pundits always mention 3 million jobs lost under Bush (a little upward rounding there), but are rarely challenged with the fact that according to the Household Survey the economy has produced 2.2 million jobs so far this year.

Some have written off the Household Survey as less reliable than the Establishment Survey. If that is the case, though, then let's dispense with the whole business of reporting unemployment rates, because the unemployment rate is determined only from the Household Survey. Further, the survey data comes from a sample of 60,000 households. If that is not a big enough sample then every political poll reported on the news is worthless (OK, bad example), given that those polls use only a few hundred or a few thousand voters.

So why are the two surveys in such disagreement? Perhaps, with established companies cautious about hiring and information technology making self-employment easier and more efficient, more people are chosing self-employment or employment at small firms.

According to this WSJ article,

'[Kenneth] Safian, who is president of Safian Investment Research Inc., based in White Plains, N.Y., says the upshot of the latest trend is that more workers are striking out on their own and earning money doing it. The economy, he says, "is becoming more entrepreneurial."'

Safian's data, from government reports on proprietor's income, indicate that self-employment is indeed rising, and showing more increases in earnings than traditional employment.

Meanwhile, the bond market is tearing higher today, which could mean that the markets think the economy is slowing. But if that is the case why is the stock market down less than 1%?

Perhaps the bond market is rallying because political pressure on the FED will make it hard to raise interest rates any time soon. That would certainly explain why bonds AND gold are higher today.

In any case, the today's news is, at a minimum, complex. My own bias is to believe this news is net somewhat good. But I freely admit it is a bias, not a complete analysis.

posted by Pete Harrigan | 11:55 AM

Thursday, November 20, 2003  

OK, time for me to respond to the next part of Karsten's post on the market and investor complacency.

First, in response to this:

All I wanted to do is get some "pure" sentiment data on the table. I don't think that anyone should rely on any one single indicator - please consider the two aforementioned ones as a backdrop for the other cited indicators. I'd like to stress the fact that I was trying to show that there is a whole series of little mosaic stones which - taken stone for stone - can all be singled out for criticism. If - in contrast - you look at "the big picture" you get an impression of a very sentiment-driven market. This can be a positive if the real world catches up to animal spirits. The way you read the data depends on your underlying assumptions about the economy - investors are either smart or greedy right now.

Yes, all these indicators do show a high level of investor bullishness right now. I will not disagree with that, because, um, it's true. My argument is that sentiment measures do not have a good track record as forward looking indicators. If you look back at a chart, you can often see that sentiment was bullish at a peak or bearish at a low point. But that is not useful, especially if the indicator turned bearish several years before the actual peak. If you can actually use an indicator to make a good trading decision, then it is worth discussing.

Karsten points to the discounts on closed end funds to their net asset value and, helpfully, links to this chart. I find this chart more useful to my side of the argument, though. At what point was the discount the largest and therefore the most bullish? It looks like about May of 2000 to me. Call me a greedy bastard, but I don't like an indicator that flashes its most bullish signal at within spittin' distance of the all time high. Mostly, this and other sentiment indicators look more like a random walk to me.

As to turnover, I'll take Karsten's word for it that is high. The table to which he links does not give a long enough time series for me to comment. And maybe insider selling is a good indicator now. I simply do not know. I'll just concede this one.

Finally, we come to PE ratios again. As to the changes in tax rates, Karsten replies that as companies are not paying significant dividends, the rate of dividend taxation is not relevant. This is somewhat simplistic. First, companies have been increasing the rate of dividend payouts since the change in the tax. Second, I would assume that those retained earnings would get paid out eventually in dividends or share repurchases, and given that the tax rate on both dividends and capital gains dropped, I would say that any given dollar of corporate earnings is worth more to the investor than it was before.

As to Karsten's very solid point that dropping inflation should appear in both the numerator and in the denominator, I agree. But I think it gets a little more complicated here. First, while lower future inflation should lower future nominal cash flows, shouldn't we still discount the current level of earnings at a lower rate if interest rates are lower. The current estimate of earnings for the S&P index (the base on which that future growth is compounded) does not drop, just the nominal level of future growth. Second, why are inflation projections lower? If it is in part due to rising productivity, perhaps the growth of future cash flows will not be lowered. Finally, why not just use real interest rates? They are certainly lower now than at any time since the wacky inflation of the late 70s - early 80s.

On the markets multiple expansion, Karsten writes:

My read on P/E's is as following: P/E's tell me what the market is expecting. A look at today's P/E of around 30 (trailing) or 18 (12 mth) tells me that investors are looking for very strong earnings growth looking forward. I just can't find any reason for this ongoing multiple expansion. The tax argument might be valid, but I just can't see it accounting for this dramatic increase in the multiple. Again: I might be very wrong - the real world could quickly catch up to the market by showing us a very sharp and sustained recovery - my problem is that I just can't seem to find evidence for the sustained part.

Well, I'll give you a reason for the sustained part. Tax rates on earnings and investment just dropped. Therefore, we should get more earnings and investment. Keynesians will assume the recovery will drop back to merely tepid after the initial effects of the tax cuts are passed. Supply siders, on the other hand, will argue that as long as marginal rates stay low, we should get above average growth rates. (Sorry for the massive simplification there.) The market seems to be in agreement with the supply siders.

Note: I hate to leave the question of Nikkei valuations unanswered, but I simply cannot find any decent data. Maybe correlations do break down under a truly terrible deflationary scenario. Meanwhile, with all the money the Fed is pumping in, that seems far from happening here, although, I understand Karsten's point was more sophisticated than that.

Note #2: Finally, you will never catch me disagreeing with Karsten on this point. Puts are cheap now. Yes, I know I said earlier that the low level of the VIX could indicate that the market is just returning to smaller swings, like back in the mid-90s. But with the VIX this low, buying some puts seems like a heads you win, tails you don't lose much kind of bet.

posted by Pete Harrigan | 1:13 PM

The horror, the horror.

Link found via GeekPress.

posted by Pete Harrigan | 8:39 AM

New jobless claims match their low of a few weeks ago. A bomb attack in Istanbul takes the S&P futures lower in the middle of the night, but now we are almost unchanged from yesterday. Looks kinda, sorta bullish from where I sit, with an emphasis on the kinda, sorta.

On the other hand, what the hell do I know?

On yet another hand (we're up to three hands now, but who's counting?), why do the news reports call 355,000 initial jobless claims signs of a "stabilizing employment situation"? Anything under 400,000 should signal expansion of the labor market. Even that 400,000 number is woefully out of date, having been calculated back when the total labor force was only 100 million.

posted by Pete Harrigan | 7:21 AM

Wednesday, November 19, 2003  

Karsten, over at Curry Blog, has responded to my criticisms of his "The Bulls They Are A Ragin" post with "The Bulls They Are A Revisited". So now the ball is back in my humble court and it is time for me to flail wildly at it. No Fear of Flairure here. Or as my mother-in-law says, "if it's not worth doing badly, it's not worth doing." Think about it (and if you figure it out, write me).

Now on to the "substance". Karsten points out that the CBOE touts its VIX index as a sentiment indicator. The index is calculated from the implied volatilities of nearby index options contracts. Thus it should indicate how investors feel about future the volatility of the market. My guess though is that the VIX reflects how the market has been behaving in the very recent past.

Therefore, in an exercise that should horrify professional statisticians everywhere, I have run a little correlation between the VIX and the Standard Deviation of the S&P 100 index (I believe the OEX options at the CBOE are based on the 100 index, not the 500 as I had originally mentioned). What I found is that there is a 76% correlation between today's VIX close and the historical vol of the last 21 trading days (roughly 30 calendar days).

This certainly proves nothing, of course, but it does fit what I have seen in the past. As a market maker at the PSE and the CME, I found that the trading floor rarely takes option volatility far from historical vol. The main exception would be the days leading up to an earnings announcement in an individual stock, but that does not pertain to an index.

I will freely admit that option trade flows determine the implied volatility in the options in any instrument, but one of the main drivers of that order flow is the past volatility. So it all gets a little circular.

My point in all this is that while the VIX index may, as the CBOE's marketing department claims, reflect the perception of market players as to future volatility and even price direction, the main reason it is lower than at any time in the last 5 years is that the market has been far less volatile recently than it has been in over 5 years.

If the low VIX is a measure of investor complacency then those investors have been right to be complacent in the last few months. At the end of the first week of January 1995, the VIX stood at 12 - 12! That makes the current reading of 18.67 look stratospheric in comparison. The market was up well over 30% in the following year. Generally, the VIX is useless as a forward looking contrary indicator.

Unfortunately, I will have to stop now and atta... address the other points later. Parent-teacher conferences wait for no man.

posted by Pete Harrigan | 2:14 PM

Monday, November 17, 2003  

Been a little busy lately. Clearly, I have had no desire to blog. Today, however, I saw this post over on Curry Blog. Now, I hate to be disagreeable, especially given that I am fast becoming a fan of Curry Blog, but I am must, however respectfully, register my objections.

Karsten makes the case that the market could be due for a crash in three ways: First, investor sentiment is too bullish (he uses several different measures for this). Second, valuations are too high. Third, really good song lyrics.

On point three, he is unassailable. I will not attempt song lyrics for fear of a truly Louis Rukeyserish moment.

Point one, however, is weaker than it seems. Yes, investor confidence is high by some measures. The first measure Karsten brings up is the CBOE's volatility index, which he points out is extremely low. Now, anyone who mentions the VIX index in a blog is OK in my book. However, the idea that the current low level of the VIX indicates investors are complacent strikes me as a misinterpretation.

The VIX index is not driven only by customer purchases and sales of options but also by the actual volatility of the market. You could buy options, thinking they are "cheap" and eat a lot of decay if the market does not move enough. Over the past six months, the the realized volatility of the S&P 500 index has been lower than at any time since a brief period in 1998, and before that 1996. The current "low" reading for the VIX would have looked "high" in the mid 1990s. The market may simply be adjusting to the end of the bubble-bear market period of the last 5 years. Perhaps investors not jumping in and out of the market can be seen as a sign of complacency, but that seems a bit of a stretch.

Karsten points out that the Yale School of Management 1-year investor confidence index is at a high level. Given that the chart indicates a certain lack of data going back past a few years, I am not sure how good a contrary indicator this is. The institutional confidence indicator has more data, and that seems to indicate confidence was almost this high in 1991 (good time to buy), 1998 (good time to buy), and 2001 (veeery baaad to buy). So, again, I fail to see how this is a reliable contrary indicator.

We also get the AAII (American Assoc. of Indiv. Investors?) Investor Sentiment Index. This is above the bearish 70 level. Of course, it went to that bearish level in early 1995 when the S&P index was around 500. Personally, I wouldn't feel too cocky having missed the bear market by means of missing the preceding bull.

All these sentiment indicators miss one major point. They point out what people say, not what they do. You can be bullish but not long, or bullish and long. Those two possibilities have different implications for the future of the market. Further, I am unaware of any time series study that has ever shown these indicators to work. That may speak more to my ignorance, though. (To his credit, Karsten also mentions the put-call ratio. Given that this is an indication of actual bets people have made, it seems a better indicator than the others.)

Finally, Karsten discusses PE ratios. This may be the time that PE ratios give a good signal about the future, but it hasn't happened much in the past. There is simply no evidence PEs work as a predictor. They were plenty high in 1932, after the market had dropped to its Great Depression low. A high PE could have shaken you out of the market in the late 90s or the early 90s. Besides, instead of just looking at past averages of PEs, shouldn't we look at the ratio as adjusted for taxes and interest rates? We have the lowest interest rates since the 1950s, and then the marginal tax on capital gains was 50% and the top tax on income (for instance, dividends) was about 90%. So while today's PE may look high, it may well be closer to the new norm ("permanent high plateau", anyone?).

I certainly don't know which way the market will go. I have been bullish since a too early July 2002. The day after the market low last October, I wrote "the market here is the mirror image of the top". I wish I had just said "Buy" and then I could claim to have made a great call, rather than just the murky statement I made. With the FED pumping in money and unable to raise rates soon, I continue to be bullish. This may look awfully silly if the last leg of the bear market is waiting to smack down on all our heads, but I figure life without the prospect of looking foolish is not worth living.

For a burst of sunny optimism, try this article from Donald "critic, not stalker" Luskin. Luskin is bullish. He uses historical averages. And he mentions Soros' "reflexivity" concept.

Reflexivity, the VIX index, all in one post. What a day to return to blogging.

(All the links related to the Curry Blog post were found through, well, Curry Blog.)

UPDATE: I completely forgot to mention where I am in agreement with Karsten. Whether or not low levels on the VIX index are a directional indicator, they definitely spell an opportunity to buy options at, well, a lower level. In one of his earlier posts, the CurryMan points out the silliness of the maxim that the majority of options expire worthless. Who cares? If I buy puts that I think only have a 5% chance of being in the money, and they expire worthless because the market rallied (oh, and I'm long, BTW), the options still had a value. They allowed me to be longer than I might otherwise be. Or if I just bought those puts to take a shot on a crash, then they still had value, because I lost less than I would have going short. Expiring worthless is no big deal.

In any case, it seems reasonable to say buying options is a decent bet, now. Either we really are returning to the low volatility of the mid 90s, in which case current vols are slightly above average, or we are just taking a respite from the high vols of 98-02, in which case options are very cheap.

posted by Pete Harrigan | 10:34 PM

Tuesday, July 01, 2003  

You have to give credit where it's due (well, you don't have to, but you should), so I must tip my hat to Nicholas Kristof, who suggested today that New York should end rent control. Imagine seeing an Op-Ed piece like that in the New York Times. And from Kristof no less. Here is what A. A. Milne calls the Honey Quote:

Right now, rents for unregulated apartments are high because they make up only one-third of the market, so they are bid up to artificial levels. If prices were freed, then retirees who spent most of their time in Florida would give up their artificially cheap three-bedroom apartments, and there would be a surge of both vacancies and new construction. Because of regulations, the average rent in New York City is only $706, and that would rise, but the real rents that people actually pay when they find new apartments would fall.

One piece of evidence Kristof did not mention, but which strengthens his argument, is that when Santa Monica ended rent control, rents dropped.

All in all, it is a great piece, if utterly shocking to this right-winger, and you should rush off and read the whole thing.

Now, if we could just end rent control here in San Francisco. Sorry, I drifted off into a dream there.

posted by Pete Harrigan | 10:31 AM