Scaling It Down
Yesterday I looked at the upper end of scalability, today I take a simple look at the lower end, again with my “strategy X.”
In the hands of a fund using all qualifiers, it holds 344 stocks on average, with 18% monthly turnover. Let’s say that, when holding only the top 20 stocks sorted by characteristic Y, “Joe Schmuckatelli, Retail Trader” experiences 25% monthly turnover in his account.
That’s 5 round trips a month, or 120 trades a year. If “Joe Schmuckatelli, Retail Trader” pays $10 per trade, that’s a $1,200 commission drag annually. Interesting, but useless out of context. More information is needed, such as the size of “Joe Schmuckatelli, Retail Trader’s” account, his tax rate, and the performance of the system.
Let’s say the system generates 15% average compounded annually – and that Joe could get 10% with very low volatility from a simple diversified buy+hold indexing across several asset classes.
Since tenure in this system averages under a year, let’s say “Joe” pays 25% of net in taxes.
Now let’s assume “Joe” has a $10,000 account – he would make $1500, pay $1200 in commissions, and be well behind buy+hold, even pretax.
With a $50,000 account, “Joe” makes $7500, pays $1200 in commissions and $1575 in taxes on the short-term gains, netting $4725, which is just below taxless buy+hold.
With a $100,000 account, “Joe” makes $15,000, pays $1200 in commissions and $3450 in taxes, netting $10,350 – just above what he could get from a taxless buy+hold.
There are lots of moving parts, not the least of which is that buy+hold isn’t taxless or commission-less; tax is deferred in some accounts, and is at the long-term capital gains rate, and there would be some expense for the reinvestment of dividends and initial buy-in. Also, one might take umbrage (I believe incorrectly so) with a low-volatility 10% average annual gain for diversified indexing across several asset classes. Finally, one can do better than $10 per trade on 120 trades/year.
The key points about scalability are: there are systems that can’t handle a lot of money, there are funds whose choices of method are hampered by their assets, and even retail schlumps need to pay attention to the scalability (down!) of their methods.


August 15th, 2008 at 4:00 pm
There is also short term tax, these indexes get rebalanced all the time and there is buying and selling in conjunction with buy-ins and redemptions that generate short term taxes.
Also the indexes are not expense free. The cheapest index fund, the Vanguard S&P 500, charges 0.18% annually, that’s $180 on that $100k account. If you are diversified across asset classes then you bet your life that the expenses of the other index funds are much higher. If you are paying an average of 0.6% in expenses (still below industry average) then you are paying half the commissions of the “active system”.
More so, many fund companies, including Vanguard, are imposing all kinds of silly fees at purchase and redemption times that are “paid directly to the fund” and thus are not counted towards fund expense ratio but do add to your cost.
Also, with Zecco ($0 for 10 trades/month), Sharebuilder ($4 per trade), TradeStation ($1 per 100 shares), ScotTrade ($7 per trade, no other fees), you do not need to put up with the $10 per trade.
August 16th, 2008 at 7:11 am
There are hundreds of ETFs with expense ratios below 0.25 - and they are primarily the index-trackers necessary for a globally diversified, low-risk buy+hold strategy. The international stock ETFs are more expensive than the S&P trackers, but the fixed-income and REIT ETFs are cheap. I think you can make it for 0.30 or less across the portfolio.
Since that strategy would be transacting once/year and holding 5-10 ETFs, the trading expense is basically “free.”
All the gains could be long-term with all the losses short-term, making it tax-efficient.
In short, it’s ridiculously easy to beat the S&P 500 long-term on a volatility-adjusted based, and match it on an absolute basis, which is why traders should pay attention to scaling their ideas down … if they can’t post a decent return, they should consider whether the effort they’re expending is worth it, or if they’re doing the work for entertainment purposes.
August 16th, 2008 at 7:57 am
Some good important points Bill - I would add:
1. I think traders often trade out of the wrong account. I have a cash account and an IRA. For my most active strategies, I actually trade them out of my IRA, whereas I basically buy and hold in my cash account because taxes can kill you. I think most traders find this counter-intuitive for most traders.
2. I am continually amazed that I find people that I would classify as traders that are still paying $10-$15 a trade. Interactive Brokers, MBTrading and Tradestation all offer 0.01 cents per share with a minimum of a $1 - so there is no reason way a trader shouldn’t be taking advantage of this. On my website, I pointed out the impact of commissions on an active strategy: http://skillanalytics.wordpress.com/2008/07/03/think-commissions-dont-matter/
3. I think you provide, without a doubt, that unless you are making sick returns with less than $10k you should be buying and holding in cash accounts.
Great post!
August 16th, 2008 at 8:07 am
“these indexes get rebalanced all the time and there is buying and selling in conjunction with buy-ins and redemptions that generate short term taxes.”
As far as I know, ETFs do not have this issue - giving them a significant tax advantage.
August 16th, 2008 at 8:27 am
I’d be inclined to a different, less U.S.-centric weighting - perhaps 1/6th in each except for the European and Asian, which should add to 1/6th total (1/12th in each) - but the basic point is the same, it can be done really really cheaply.
http://www.indexuniverse.com/component/content/article/31/3426.html?Itemid=3
August 16th, 2008 at 11:12 am
““these indexes get rebalanced all the time and there is buying and selling in conjunction with buy-ins and redemptions that generate short term taxes.”
As far as I know, ETFs do not have this issue - giving them a significant tax advantage.”
My point is that the underlying index, the S&P 500 for example, changes components several times a year as the S&P committee adds/removes companies from the index. Also companies get bought out, merged, bankrupt, delisted, etc… throughout the year. Index Funds and Index ETFs have to buy/sell to reflect those changes. Most people forget that the Broad Indexes are actively managed by committees at the S&P and Russel companies. Both Index Funds and ETFs have to match these changes.
These transactions create drag both in terms of transaction costs and in short term taxes, whether it is paid by the end-user or the fund itself, the cost is paid by somebody.
Index Mutual funds also have to keep cash on hand for redemption so they are not 100% invested. ETF’s have other guidelines on how/when they create new blocks of shares that also creates drag.
My point is that commissions at this day and age should not matter unless you are an active day trader. If you are a system trader with 10 or so trades a month then the difference between buy-and-hold of Index/ETF funds and your buy/sell of stocks is negligible in terms of commissions because the underlying ETF/Index funds have expenses/drag on par with what you pay for commission.
August 17th, 2008 at 3:11 pm
Bill,
Have you accounted for the trading slippage costs (the difference between the bid/ask spread) in your system results? This expense has the potential to negatively impact net returns to a far greater degree than trading commissions.
August 17th, 2008 at 5:42 pm
No slippage is included. I don’t believe that slippage, for *these* systems at retail size, is nearly as significant as commission expense. For slippage to be a negative impact would be to assume that the movement from signal to execution was, on average, adverse to the system, and was additionally rather large. This makes more sense when the system is keying on a fast-moving market intraday and watching for signals on a constant basis; when the system is making changes at open based on regularly-scheduled looks at closing prices once every so many weeks … I don’t think slippage is as large a consideration as transaction expense.
In the previous posts on scaling systems UP, slippage is the primary consideration in how big a fund can get, how much liquidity it needs for its trading issues, before it moves the market. Slippage is a lot different problem when you’re causing it yourself …