![]() | ||||||
| ||||||
| ||||||
|
We define ultra
small as companies the size of the smallest 10% on the New York Stock
Exchange, which is to say currently anything less than about $140
million in market capitalization. The Center for Research in Security
Prices out of the University of Chicago takes the NYSE and ranks it from
top to bottom by market cap and puts them in 10 equal baskets of stocks,
about 160 or so stocks in each basket. That sets the size range that
they call deciles. The 10th decile is what we call ultra small and the
9th decile we call micro caps. How so? Part of the difference is that if you look over the last 7 1/2 decades, 10th decile, or ultra small stocks, have returned an average 13.1% annual return over that full period. The average small-cap stock has returned about 11 percentage points. So it's more than a 2 percentage point difference and most of that differential is just from the 9th decile to the 10th. A 2% edge a year over your average small-cap fund is a lot of money in the long term. If you take a 30-year period of time and you have 2% more a year, you're looking at 80% more money after 30 years. If it's somebody's retirement money and they have 30 years to go, this asset class based on longer-term numbers historically would get you an 80% higher retirement. How does your fund exploit this universe? What is your strategy? Our strategy is to
tap the financial characteristics of a unique asset class. There are
about 2000 stocks or so that are in the universe of ultra-small stocks
that we track, and we have two portfolios. One of them, our Ultra-Small
Company Fund, is a very active style. We run our quantitative models to
beat the market and buy the best stocks we can with that fund, which
owns on average about 150-160 positions. The fund has an expense ratio
of about 1 1/2 percent, which is pretty lean by industry standards of
even micro cap, but is higher than our other ultra-small
fund. What ways are those? One is on the trading cost. The bid-ask spread on stocks this small is on the order of 4-5%, so you can't just go out and tell your broker to put a market order to buy 100,000 shares or 1000 shares for that matter. If the spread's 5%, let's say, and the true market is half way in between, then every time you bought a stock you'd be giving up 2 1/2 % in transaction cost . That's about what you expect the asset class to outperform over time. So trading these stocks is very time-intensive and it's one of the reasons we think no one else offers a stock portfolio of stocks this small. We devote a lot of resources to the trading desk here at Bridgeway. Out of six people in our investment management team half of those are in the trading area. Our goal in our Tax Advantage portfolio is to buy a stock on average over time a little closer to the bid than the ask, which is a neat trick if you can pull it off. I think of it as the "frictionless wheel." To have a net no trading cost is a huge advantage over even the average small-cap fund, where you're going to pay upwards of 2% in trading costs every time you buy or sell a stock. If you have an actively managed portfolio with 100% turnover, you're paying 2% to buy a stock and 2% to sell a stock. That's a 4 percentage-point hurdle every year, and we're trying to get that down in the Tax Advantage portfolio essentially to 0. On top of that we're in an asset class that has a couple percentage point advantage, and on top of that our expense ratio, instead of being, let's say, 1 1/2 percent, is 75 basis points, so you save almost a percent there. I'm looking at a 5 percentage point advantage with Ultra-Small Company Tax Advantage over an actively managed small-cap fund before we get out and even buy any stocks. Aren't you also trying to save on trading costs in the Ultra-Small Company Fund? We try, but that's a more actively managed version, where we go out and buy those stocks faster. We're only going after the cream of the crop in that portfolio, and those stocks tend to move faster, so we're going to ante up a little more to get those companies right away. 500 stocks, or even 150, is quite a large number. Why so many? It's enough to reasonably diversity away the company risk, and it does take a lot more diversification when investing in companies this size. One way to think of it is that in any one year there are going to be a lot more ultra-small stocks that can double and quadruple. If you're a tiny company that owns .01% of the market for a given product, let's say a soft drink company and we've owned one, and you wake up next year and you own .05% of the market, your revenues are up five-fold and Coca Cola still doesn't know you exist. Microsoft and GE aren't doubling revenues next year, but a reasonable percentage of ultra-small socks can go up strongly. The downside, or reverse, of that is that a much higher percentage of ultra-small stocks go completely out of business every year. Don't you have to do far more research on these companies and their management given these risks? That observation is the reason no one else offers a stock portfolio of ultra-small companies. In fact, we're a pure "quant" shop. All of our modeling and understanding of these companies is based on numbers and statistics, mostly fundamental information that gets filed on Edgar: financial statements, income statements, balance sheets, cash flow statements, and then there's some technical data that goes into some of our models. So the good thing is we don't have to go out and interview management. We don't do any market timing across the board at Bridgeway. We don't do any forecasting of the economy. It's all a bottom-up, stock-picking, fully invested methodology, and that's true for all of our funds. Can you be more specific on your quantitative screens? We don't talk about the specific inputs to the models at Bridgeway, so I can't talk to you more specifically. It sounds like you're protecting the Coke formula. (Laughter.) That's exactly how my brother describes it. What style area do your ultra-small funds fall into? We use models that span the style box, so it's much harder to peg. Both ultra-small company portfolios tend to show up on the more value-y end of the spectrum because the whole asset class tends to have lower P/E's, price to book, and things like that, but they've done so well in the last few years that I think we've slipped over into the blend category, according to Morningstar. It's not that we're doing anything different or we think all of a sudden growth is a good thing. It's just where those 2,000 companies show up. 1999 was your best year with 40+ percent returns, which seems odd for a value- oriented fund. There are some ultra-small stocks in technology, and we did have some of that, and in fact we made a pile of money off a few Internet names back in the heyday of the Internet boom. What industries come up currently in your screens? Banks have been pretty strong in our ultra-small portfolio. Some of the home building stocks have been there, and related ones like furniture are interesting. The one that's probably strongest out of proportion with normal is telecom. What we've seen there is that telecom stocks both large and ultra-small were just trounced in the 2000-2001 period and there was a fair amount of red ink around. The valuations got very cheap over the last couple years, and the fundamentals started coming back in the last half year to year. Our model completely takes the emotion out of the process, and in a turnaround situation Wall Street tends to have a wait-and-see, prove-to-me-if-this-is-real attitude, and our models don't have that, so we've been more strongly invested in telecom. Sometimes the models bring up companies that I think are great, and other times it brings up ones I think I would never invest in if I were doing this the classical way of thinking about the economy and good companies and what's likely to happen. For example, Ask Jeeves (ASKJ) is one our best performing in our Tax Advantage portfolio. It's a little dot-com, and after all the carnage the Internet business took people through in 2001 and 2002, that's probably the last place I would have thought to invest in stocks again. I don't think I would have picked up on that one myself, but we own a number of Internet stocks, and, quite frankly, a large number of them have become ultra small. Any stocks in particular come up highly rated? One company we've owned and that has done extremely well for us is Clayton Williams (CWEI), an oil and gas firm in Louisiana and Texas and some other states. Sometimes it's hard to see why we hold the stock because they're all multi-factor models. You put two companies side by side that have similar growth rates and valuations and the technicals look similar, and it's just a confluence of factors in the model that make one a buy and one not. On this particular one, they had a blow-out quarter in March. That's one of the things. They had earnings of $1.71 in the March quarter and estimate 19 cents here in the next quarter after a bunch of red ink last year, so this is a company where the stock price was pretty cheap. It's taken off like a rocket and interestingly the valuations aren't necessarily way out of bounds. Cash flow on this company looks good. Our models look at fundamental data, and this particular one is a kind of "growth at a reasonable price" model. |
| |||||
|
Ultra Small Company Top 10 Holdings as of March 31, 2003: 1. FTI
Consulting How did you come to use quantitative modeling? I worked with computer modeling and quantitative methods as a research engineer at MIT in the late 1970s. I took two years off and went to Harvard Business School, and in one of the courses we had a case study on a quantitative firm. It was a firm with a very attractive track record, and at the end the professor says, "Now how many people think you can get out of business school and beat this record?" Quantitative analysis is just numbers and statistics and not very glamorous, and a lot of people want to think they're the next Peter Lynch - that is, hold the product in your hand and scratch your bead and ask the company executive some probing question, and quantitative modeling is not that way. I like numbers, but quite frankly most people in life find this stuff quite boring. So we were in this class of bright young men and women, and 80% raised their hands. Now at least 80% -- and I think it's upwards of 90% -- of money managers under-perform the market benchmarks over long periods of time. So you've got 80% of the people raising their hands saying they're going to beat it when you know in fact it's probably less than 20% and probably substantially less than that. So you saw an opportunity? That's right.
Clearly, there's a strong emotional tie and optimism to thinking you can
beat the market that just perpetuates itself. If people believed the
numbers on active management, nobody would hire an active manger,
including me. What I really got out of that class is there's an
opportunity when the people that are in the positions to make the
decision think they can beat it when in fact you know in aggregate that
can't be true. That was my insight from business school, and I thought
there should be an application for quantitative modeling. I thought
essentially what quantitative modeling would do is take emotion out of
the process, and that's the biggest destroyer of value in our industry.
| ||||||