There was an interesting flurry of comments yesterday regarding sector weightings. I left a chart up from Bespoke and made the comment that I believe in staying in touch with sector weights and to lighten up when a sector gets out of its historical whack.There were two comments by one reader saying that it makes more sense to "to compare price to book of various industries with their historical norms and bet on some reversion to the mean."
In a subsequent comment he says "it doesn't make any economic sense to talk about what is the correct percentage of market cap weight for 'tech' in the S&P 500" and then repeats his preference to price to book.
Another reader sort of called him out essentially saying there is more than one way to skin a cat.
I view this whole thing much differently than the commenter. I view the price to book belief as more of a bottom up approach. I will not say it makes no sense but I do believe bottom up is the more difficult method.
In the early 1980s energy comprised almost 30% (it may have been 26% actually) of the S&P 500 and then imploded in on itself to a larger degree than tech did at the start of this decade. These two examples are extreme and they don't happen in this magnitude very often. The financial sector is currently the largest in the S&P 500 and has been over 20% of the market for quite a while. This has been part of my thinking in not wanting to go heavy (in addition to concerns about the slope of the yield curve).
Managing this sort of thing is very subtle and a bottom up person probably doesn't care but I think it can really help out every now and then.
As far as the comment on tech, I view this differently too. There is precedent for semiconductors to provide leadership early in the cycle and for software to do a little better later in the cycle. It might be worth knowing that before you load up on semiconductors at the end of the cycle. Buying this sub-sector at the wrong time creates a headwind. You might still do very well but you are overcoming a headwind to do so.
Using an example from the other day with financials. I talked about adding a little volatility in the sector once the yield curve gets right side up. Inverted curve, late in the cycle calls for, IMO, one type of exposure. A steep curve early in the cycle calls for another type of exposure.
This carries over to various sectors. Also at different points you want a different average cap size in the portfolio or tilting to growth or value is the better place to be. Again this is subtle stuff but I am convinced that I have added basis points making these sorts of decisions in client portfolios. This really is what top down is, it determines a lot of the return to be had and I think requires being less right than owning a bunch stocks that should go up because they are cheap.
Carl Yastrzemski might be the least talked about hall of famer from the modern era, outside of Boston anyway. You can read more about him here.
17 comments:
Mr. Randomness, could you provide comments on this article relating to current discounts in CEF's please.
http://www.thestreet.com/s/12-dividend-yield-whats-not-to-love/funds/mutualfundinvesting/10375554.html
and
http://www.thestreet.com/s/make-your-move-on-cheap-closed-end-funds/funds/mutualfundinvesting/10375199.html
Roger
Is that your personal copy of a Yastrzemski card?
Is it not sad how Topps and other companies drove the baseball card industry into the ground? I recently sold my entire collection which included many rookies for $80. I paid probably a few thousand for my collection back in the day.
The Baseball card industry reminds me of the US Dollar. You'd think our Federal Reserve would take a queue from Topps. If you print too much of something, its not worth crap.
the 12% looks good of course and while one year from now you might be glad you bought now, a month from now you may not. if there is another leg to this em will likely get hit again.my CEF exposure is not that heavy because they do get pasted in these types of events.
Not my Yaz card. I had a collection when I was highschool. I left them at my mom's house when I left after highschool. She moved a year or so later, wrote Roger's card on the boxes and they disappeared. Very typical story, I had some good ones.
Much like emerging markets where once the foreign money leaves the locals are left holding the bag...I just bought a Yaz non rookie for $80 bucks in Beantown...sigh.
Sector weights in the economy do change over time. Maybe some sort of moving average approach could work? Lighten up on a sector when it gets to far from its moving average. But you'd need monthly data or something for that.
Our price to book friend is only looking at one aspect of a stock value, thus he missed Roger's point. A company's stock value is somewhat determined by their earnings; but also in the short term it can be weighed down or up by what sector it is in.
For instance if a whole sector gets crushed like the tech sector did in 2000, even the good companies go down with the bad. By the same token a lousy company can get an artificial lift if the sector they are in gets a bounce based on some good news that a good company gets in their sector.
John Dorfman is a big believer in looking at the low price to book and low P/E of a company when considering a purchase of stock though. I'm a fan of Dorfman's column myself. Has anyone heard of his "Robot Portfolio"? It does quite well buying with this method of choosing cheap, out of favor stocks and holding them for a year.
http://www.bloomberg.com/apps/news?pid=10000039&sid=aDfZA.kxtYKk&refer=columnist_dorfman
An interesting read here by Jeffrey Cooper on fear and the upcoming lunar eclipse. (I assume he was joking about the eclipse part...but maybe not)
http://www.minyanville.com/articles/Paulson-quant-fear-risk-fundamentals/index/a/13836/from/yahoo
I found it interesting that he says that fear usually lasts a minimum of 55-90 days.
There are several different points of disagreement and confusion here. I'll tackle them all here, but try to keep them separate.
Disagreement 1 - Is betting on mean reversion of broad sector weightings a good investment strategy? I say "no" - both because it doesn't make economic sense, in an of itself, and because it is easy to replace it with more rational strategies that apparently accomplish the same goals (improving portfolio performance with mean reverting re-weightings). So I am not claiming that betting on reversion of sector weightings will prove harmful. Rather I'm saying that it doesn't belong in one's arsenal of top down considerations - i.e. to the extent one wants to use top down criteria, sector weights reversion should be replaced with more sensible top down criteria.
As an alternative, I offered the example of betting instead on some mean reversion of average price to book in capital intensive industries. This was not a suggestion to use bottom up criteria instead (it's true that I think bottom up criteria should carry more weight, but that's a debate for another posting and wasn't part of my previous criticism). I didn't specify details like how to calculate historical P/B for industries - to be concrete, I'd probably prefer revenue based weightings for industry constituents for a given industry in year X1, and let's assume we are weighting year X1 vs. year X2 in the same way one would do it for historical sector weightings. Maybe Buffett is betting on railroads/transports because they had a huge sector weighting in the 19th Century? :)
One of the ideas behind looking at price to book is that it factors in the effects of increasing or decreasing competition - since book value is related to the cost of generating competition, industries with high price to book tend to attract a lot of new competition and industries with low price to book tend to discourage new or marginal entrants.
Since book value is also closely related to retained earnings, P/B also tends indirectly to say something about how future optimism compares to past financial performance; since future optimism (or pessimism) often swings too far, some mean reversion there also makes sense.
Hopefully the notes above clarify both that one can use P/B as part of a top down mean reversion strategy and that there is at least some semblance of logic behind the idea. Why do I say this should this be better than a strategy based on sector weighting? One reason is that the concepts of sectors are so vague and detached from economic realities that economic generalizations don't apply well. For example I criticized the concept of "tech" because it's apparently based on an incidental example of business implementation "business has something (perhaps indirect) to do with a semiconductor, possibly just because a computer or a network is involved somehow". So the market wants to include not only chip makers and chip maker suppliers but also internet retailers and advertisers and telecom companies and software companies but not usually medical technology companies or chemical engineering companies, etc. This is not a sensible economic equivocation. Some of these companies are capital intensive and some are not. Some benefit from IP protections and some are hurt by it. Some benefit from excess fab capacity for semiconductors and others are hurt by it. Some products of scientific/engineering innovation are local to "tech" and many others relate mainly to other sectors (e.g. bio in healthcare, chem in materials). Tech is just not a useful category for talking about fundamentals of business markets and cycles. Roger responds above by talking about the business cycle and "semiconductors" - if we unpack the ideas behind what he is saying, it will turn out to be a reasonable claim about a few specific industries, not about "tech", in the sense of a broad market sector.
Confusion 2+3 - Is the disagreement is based on bottom up vs. top down? Is using P/B my preferred investment method?
As discussed above, the P/B alternative was meant as an apples-to-apples top down alternative. It was not meant to express my personal investment methods. It's true that I am more of a bottom-up investor, but I'm not arguing for that here and P/B doesn't typically play a central role in my selection unless I am looking at an asset play (it's much too simplistic and doesn't take into account forward growth prospects for an industry or a stock).
Confusion 4 - Am I denying that sector based thinking is important to many market participants?
No I'm not. If you look back to my original post, I said "Am I the only one in the universe who thinks..." Rather I think that sector based thinking, to the extent it overrides more fundamental considerations, is irrational, so I bet on it not overriding other factors when predicting long term stock performance. It's definitely true that sometimes this can bite me. For example, at end of June, early July, I noticed that many insurers had very low valuations and that their earnings estimates as a group were going up. At the same time I was negative on the banks/brokers and the home industry. Well the insurers did report very good earnings that even beat estimates as a group, didn't have much or any sub-prime or CDO exposure (especially the small ones that interested me), and they got temporarily crushed in late July because people were shorting the "financials". Obviously if I could have predicted that the panic would hit "financials" in late July, I would have avoided this industry. Now they seem to have bottomed, and in recent weeks the insurers are outperforming. So it goes...
Good post Josh. I agree that some sector definitions leave much to be desired, particularly ones such as 'tech' which I tend to subdivide on fundamental grounds in any case when evaluating assets for inclusion in my strategic portfolio. Your last point however illustrates why I normally don't do that WRT my tactical portfolio because trading there is shorter term and typically more dependent upon the perceptions of other investors rather than fundamentals alone; e.g., if most investors consider 'financials' the same they are likely to be unkind to banks and insurers alike in a downturn.
Assuming reversion to the mean is reasonable in terms of weighting regardless but I'd advise readers to keep in mind that reversion probably does not reflect any deep economic principle or a fundamental quality of the markets. That is, reversion to the mean is almost certainly a 'simple' statistical outcome of imprecision in measurement so it may not pay to assume it is inevitable; it's such a common phenomenon though that this is more of a pet peeve than anything else.
Agree that reversion to the mean is always just a heuristic. I'm not arguing for emphasis on such heuristics as a primary investment factor. Rather, I'm saying that if one is using them as a secondary investment factor (and I think most all top down theories fall in the category of secondary investment factor), then some instances of this form of heuristic will have more predictive power than other instances - because they capture more fundamental aspects of the business environment. I'll use *my predictions* about short term market factors if I have a basis for being confident in those predictions. But they are a tertiary factor because it's hard to make such predictions reliably, so under typical market conditions it's hard to have confidence in them.
It' tricky to work out the interplay of primary, secondary, and tertiary factors. Here's a current example: Amrep/AXR - horrible chart/industry, it trades with the homebuilders and short term profits depend on real estate development in Rio Rancho, NM. Like most everyone, I'm still bearish on homebuilders. But this company has no debt, earns $1/share from other revenue streams (magazine distribution and investment interest), and the undeveloped land it owns is worth at least $120/share ( or at least $80/share after taxes and selling costs); long term growth prospects for Rio Rancho are very strong; and the stock is around $36. The fact that AXR's narrow float is heavily shorted makes this situation even more tricky. So from day to day, the primary, secondary, and tertiary factors may all point in different directions. I'll also note that real estate is likely to decorrelate from the market long term, but the housing factors may be highly correlated here in the short term. Very tricky...
Tricky indeed. The real estate environment is extremely risky right now and I'm not long any home builder or mortgage lender but still short several although I did close out my WCI positions this week; $21+ down to 7+ is good enough. Longer term, oh yes, I'll play but not for the next few quarters I think. Always ready to change my stance should circumstances and data warrant of course.
Headlines on CNN News: "States and nonprofits rush to provide help to those victimized by subprime loans, and borrowers rush to accept"
I could not believe this when I read it. These people who borrowed money for houses they could not afford are now considered to be "victimized"?
Headlines on CNN News: "States and nonprofits rush to provide help to those victimized by subprime loans, and borrowers rush to accept"
I could not believe this when I read it. These people who borrowed money for houses they could not afford are now considered to be "victimized"?
Here are a couple of questions for anyone. There is a relatively new ETF (PYH) that uses the Value Line Industry Rotation Schedule that apparently keeps current with sector strength changes. Anyone think this is worth looking at as a strategy?
Second, looking at Iran buying nuke-ready Russian jets, Russia thumping its chest, and China getting ready to deploy an aircraft carrier in the near future, can anyone argue that defensive funds like PPA are not good investments for the forseeable future?
Scoot
Roger, I just sent 4 charts to your Yahoo account.
some were victimized. I dumped my real-estate agent few years ago when i kept insisting on buying a house for under a 1/3 of what i qualified for and he kept showing me much more expensive houses.
He kept giving me newspaper articles about the year-over-year appreciation in the area.
He kept saying: "You live in a better house for few years and make out much better on the way out". Why would you buy for less than you qualify? he kept asking.
The appraisers jacked up the prices. The counties kept increasing the "values" of the homes to collect more taxes.
Mortgage brokers "fixed" the paperwork till they got you as high a mortgage as possible and then sold the mortgage.
some were victims, some were stupid and many were a bit of each.
FWIW: The mortgage portfolio "owners" cannot afford all the foreclosures.
The solution is actually very simple. What will happen is that most ARMs will not increase when they reset. The interest rate does not have to increase upon a reset. Few years ago some rates actually went down upon a reset.
The "smart money" that will end up buying those mortgage portfolios at bargain prices will simply reset the rates at a very slow pace and this problem will be contained.
Real-estate prices will suffer for a while though.
Sectorology.com - Watch Stock market videos
Post a Comment