| |
 July 2011 Quant at a Cost Can Investors Afford Model Portfolios P3 and P4?
Quantitative strategies -- those based on market trends and trading patterns of the past rather than strictly fundamental factors -- are often feared by investors because they tend to require a large number of trades. Some do, but that's not the case for all. When quantitative models P3 and P4 were created, holding down transaction costs was one of the goals. Over the years, however, we've actually modified their trading strategies in ways that have actually increased trading (see, Time for a Change and Old and New). Now it looks like turnover is on the increase.
As suggested above, turnover can be a real issue for investors, especially quants. Because quantitative portfolios are rebalanced on a regular schedule, there's an increased likelihood of transactions and their associated costs. Discount brokerage rates can help restrain the costs, but even the cheapest begin to add up as the number of trades increases. At some point the costs start to outweigh the benefits, so investors are wise to keep a wary eye on the required level of trading. P3 and P4 are rebalanced regularly throughout the year. P3 is revisited every two months but its trading is limited by its required number of holdings: It must always have thirty stocks, no more and no less. P4 has held as many as 68 individual issues and as few as 31. As we'll see in a minute, this has a major effect on turnover.
Understanding Turnover and Its Expense Turnover is, strictly speaking, the number of holdings replaced in a portfolio in a one-year period. For example, if you started the year with five stocks (A, B, C, D, E) and finished it with five others (F, G, H, I, J) you would have replaced each stock with a new one. In this case all of our initial holdings are now gone so your turnover is 100% (5 stocks replaced divided by 5 stocks starting). If instead, you finished the year holding stocks A, B, C, F, G) then only two were replaced and turnover would be 40% (2 replaced/5 starting). The inverse of a portfolio's turnover gives you an idea of its holding period -- how long (on average) it holds a stock before selling it. In the earlier examples, a portfolio with 100% turnover has a holding period of 1-year (1/100%). This makes sense because all holdings are replaced within the span of the year. The portfolio with 40% turnover holds its shares for two and a half years on average (1/40%). In other words, on average, all of today's holdings are all likely to be gone in two and a half years. (This is hedged a little because one or more holdings may survive longer, but others may be both purchased and sold within the period still resulting in the 40% turnover rate.)
So far, this description of turnover applies to P3 which must always have 30 holdings. When one is sold another must be purchased so trades are always one-for-one. But P4 doesn't have this requirement. Instead, when it is reoptimized it's always balanced back to the sector weightings of the benchmark, S&P 500. That's why over the past eleven years it's number of holdings has varied considerably. Obviously, trading is not always on a one-for-one basis for P4. That's actually the way it is for most portfolios including separately managed accounts and mutual funds. This is where it gets interesting. Going back to the basic definition of turnover -- the number of stocks replaced in a portfolio over the year -- it's general convention among money managers to measure it by using the minimum of buys or sells when calculating turnover. This number is divided by the starting number of holdings to determine the portfolio's turnover. For example, again assume a portfolio begins the year with five stocks: A, B, C, D, and E and finishes with A, B, C, D, F, G, and H. In this case there was one sell (E) and three buys (F, G, and H) ending the year with seven holdings. Turnover, is the lesser of the buys and sells (1 sell in this case) divided by the starting number of holdings (5), so is 20%. If you think about it, this makes sense because even though three stocks have been added, only one (E) has really been replaced. In fact, had E not been sold and F, G, and H had been added, turnover would be 0% (0 sales/5 starting holdings). While this may be an accurate statement of portfolio turnover, it systematically understates the number of transactions. When stock E was replaced in the example above, there were actually four trades, three buys and one sell, but the turnover calculation only uses the minimum from either side, in this case 1. When determining the true cost of maintaining a portfolio, it's necessary to focus on the total number of transactions because each incurs a cost. Your broker doesn't care if you have an offsetting buy for every sell or not, he's going to charge his commission on each and every trade.
The Cost So Far Transactions are even higher. From inception through December 2010, P3 had a total of 592 and P4 was almost double at 1146. Then again, those totals are for more than ten years. On an annual basis, it works out to 54 and 104 per year, respectively. That's still a lot of trading. And it may be getting greater. Chart 1 shows the annual number of trades for both models. P3 has generally had fewer transactions, but both are trending higher. Since inception, P3 has ranged from a low of 14 to a high of 110 while P4 went from 48 to 170. But ranges and trends aren't the issue, dollars and cents are. Are these quantitative portfolios cost effective? Since we have the average number of trades per year, it's possible to estimate an annual investment cost by simply multiplying by the associated commissions. Because trades are determined by a quantitative algorithm rather than research or expensive advice, there's really no reason to pay full-service brokerage costs. Instead, an online trading account through a reliable discount broker will be more than sufficient. It's not hard to find online brokers that charge $10 or less per trade, some are even cheaper. Since $10 is such a nice easy number to work with, let's go with that.
At $10 a pop, P3 would run an investor $540 a year, while P4 would be $1,040. Is that too expensive for you?
A Real-World Comparison The math is simple, all by dividing the annual cost by the funds' average expense ratio, we can arrive at the breakeven portfolio size. For P3 this is $540/1.45% = $37,241. For P4 it's almost twice as much, $1040/1.45% = $71,724. If you're just looking to dabble with $5000, stick with the mutual funds. Costs for portfolios that small would be 10.8% for P3 and 20.8% for P4. That's not worth it, so for them, funds are generally the only viable alternative. On the other hand, if you're willing to use P3 and P4 for any amount over the breakeven points calculated above, you'll not only enjoy lower overall expenses, you'll also get performance that stacks up well against comparable mutual funds. You won't sacrifice active management or portfolio diversification, either. In addition, the more you invest over the breakeven point, the lower the expense. By the way, all the foregoing has been based on our two quant portfolios that focus on large cap stocks. Expenses would be even lower for the other two that utilize only exchange traded funds (ETFs). This is a little counterintuitive given that unlike individual stocks, ETFs carry expense ratios which must also be factored in when calculating total investment cost. Nevertheless, because P5 and P6 have a very limited universe of potential holdings (9 and 5, respectively), transactions are minimal, essentially just leaving the ETF charge. Even with a discount brokerage charge of $10, transactions will generally be more expensive than ETF expense ratios. The only time this won't be the case is when the portfolio is extremely large. So, yes, well constructed quantitative portfolios can be cost effective. In general, they become even more so as the size of the portfolio grows. Not all quant models can say this, particularly those that are designed for rapid-fire trading. We prefer to think of our models as investment models, not trading vehicles. In fact, that was one of the major considerations in their original design. The results reflect their success. Search this site! Just enter you key word or words: Get current quotes or follow your own custom portfolio,
courtesy of E-Line Financials:
|
||||||||||||||||||||||||||||||
T'S EASY TO CONFUSE INVESTING with trading. The talking heads on CNBC are always telling you it's time to get into or out of the market. Magazine covers scream about the "10 Stocks to Own Now", and even mutual funds tout their short-term performance. All of this is trading which, in and of itself, is not the object of investing. Instead, investors should see superior risk-adjusted return rather than activity. In fact, given that trades aren't free, anyone seeking to maximize return should attempt to minimize transactions.
