Position Sizing Q & A

Here are my responses to the comments left on last week’s “position sizing considerations” post:

nonadamas wrote:

1. What about equipment, ISP failures, do you always write down your trades?

2. What kind of backup set-up do you have?

3. Brokerage problems, problems at their site or with their people making errors?

When I started active trading, I would use huge amts of margin, and did not think of possible failure scenarios? So now I am more thinking it through, prompting these questions – and using margin more cautiously of course.

My response:

1. I’m not sure I understand the part about writing my trades down. I place stops immediately after entry. The stops are persistent on CyberTrader’s server so if I have some kind of equipment failure my stops will protect me. I usually begin logging my trades in my journal in the afternoon and I finish shortly after the market closes.
2. My cellphone is my backup. I can either call and close positions or do it via my phone’s web browser.
3. If my broker has problems I either contact them via the built in instant messaging system or over the phone.

Max wrote:

Wait Mike you’re saying at some point you would have 100% of your equity in a position? Do you find this dangerous? What happens when you run into a bad streak of trades? How will you deal with the drawdowns? The most I ever have had on a position is 25% and the stop was at 1% of that. Can you talk more about your position sizing and why your comfortable with it.

No I don’t really find it dangerous given the stipulations I stated in the previous post. I don’t mess with stocks that I think are likely to get halted and thereby have the potential crush me. Of course any stock could get halted at any time but some are much more likely than others. The beauty of day trading is that for the most part you can precisely control your risk. That assumes you’re dealing with liquid stocks and using market-held (hard) stops. If you ever wanted a reason to place real, physical stops, using a ton of leverage is it. You will eventually get killed if you size your positions the way I do (percent risk model) and don’t cut your losses.

Just because I have 100 or even 200% of my equity in one position doesn’t mean that I’d lose more than I’d lose if I only had 25% of my equity in one position. This is where risk control comes in. There’s a difference between dollars at work (position size) and dollars at risk. What at risk is how much I’ll lose if my initial stop gets hit. My risk per trade is constant regardless of the number of dollars at work. My risk per trade (R) at the moment is 0.75% of my equity. If I hit a string of losses (which I’ve certainly done) I’ll lose some multiple of 0.75% regardless of my position size. And, as my equity shrinks my dollars at risk per trade will shrink as well.

Steve wrote:

Love your blog. I think it’s brilliant and I’ve learned a ton since I found it.

You seem to have a bias for higher priced stocks I’m not sure I understand. I see the issue being more of relative market liquidity and volatility, neither of which will necessarily be dictated by price.

Commission on the $50 stock is less expensive if you are paying per share, and I do understand they add up at the end of the year, but it is a cost of business that averages out.

Seems like the tail wagging the dog to me. If commissions are less than an acceptable percentage of your gross profit then commissions are really the least of your worries, work on gettin your expectancy up.

In your example you are assuming the $5 stock purchase of 20,000 shares would have more market impact than the 2,000 share purchase at $50. But price alone doesn’t tell the story. If the $50 stock has avg daily volume of 1 mil, and the $5 stock has avg daily volume of 10 mil, then they have the exact same daily market liquidity. Why would buying 20,000 of the $5 stock be any more likely to “move” the market?

Also, you seem to be assuming the volatility characteristics of both stocks are the same and the likelihood of having (x)R return on both stocks is equal. Something that is not likely to be true and also something that is hard to determine in the heat of battle when they both pop up on the day trade radar.

So having a bias against lower priced, but highly volatile and highly liquid stocks could be keeping you out of some trades that might be considerably more profitable than their higher priced but less volatile counter parts.

Does this make sense or am I missing something fundamental in what you were trying to get across?

Thanks again for providing such a wonderful forum. Please keep on.

I probably do have a bias against low-priced stocks and, in theory, that makes no sense — just like all the people who favor low priced stocks. (Would you rather have a $50 bill or 10 $5 bills???) As you point out, the price doesn’t matter. What matters is the number of dollars at work and the percentage moves. As long as a stock is liquid the price doesn’t really matter to me. I do trade some sub $10 stocks if they present a decent setup. But there’s also plenty of volatility in the high dollar stocks. Too many people are concerned about buying X thousand shares instead of focusing on dollars and percentage moves. But that’s another topic for another post.

Yes, in my example I did assume that the volatility characteristics of both stocks were the same. I was just trying to give a simple example. I think we all know that in real life things are never so simple, and especially not so simple before you’ve even entered the trade and don’t know how the stock is going to act.

You make a good point about commissions but for me (and I can only speak for myself) per-share commissions save a ton of money. Maybe if I only focused on low priced stocks and/or grew my account so that I was trading much bigger lots I’d switch back to per-share commissions. As I wrote in May 2005:

The $9.95 commissions may not sound bad but they can add up quickly. Yesterday, for example, I would have generated $140 in commissions had I been trading in live mode. I traded six different stocks, taking partial profits in two of them mid-day. Those same trades would have cost me about $19 with per share pricing. So that’s a savings of about $120. If that’s a typical day those savings work out to about $2,400 per month or $28,800 per year.

Again, somebody else’s situation may be different, I can only speak for myself. Every trader should do the math for themselves.

Jet wrote:

Hi, great blog. I must admit that I was surprised that you are still paying per share commissions. It would seem that flat rate commissions would be the way to go. That way the impact on your risk calculations would be minimal, ie, why not virtually eliminate one bad thing (large per trade commissions) that are having a disproportionate impact on your risk by going with a flat rate commission broker. The impact on risk management when one has to take a relatively large commission into account seems unacceptable to me. Why not eliminate something that impacts negatively? With a flat rate commission structure, the impact of commissions is negligible. You can then trade whatever number of shares suits you without regard to commission costs, as one trade will cost the same low price as the other. Your commission structure seems to have a disproportionately large and negative impact on your trading. If your commissions are so large as to be a consideration, I would assert that you need to find another brokerage (say, E*trade, for example) which will allow you to almost entirely forget about commissions. One less bad thing to worry about.

Wow, the day I switch from CyberTrader to E*Trade will be the day… oh boy! (I use e*Trade for my IRA but would NEVER use them for day trading — except maybe E*Trade Pro.) I don’t know what Jet read but like I wrote above the per-share commissions work out much better for me. Perhaps I used too extreme of an example in my previous post. Just by the way I trade 99% of my lot sizes are less than 2,000 shares. A 2,000 share lot will cost me $12 each way or $24 for a buy and a sell. The vast majority of my lot sizes are under 1000 shares so.

As for me “still” paying per share commissions, I think the majority of active day traders go that route. Most (all?) flat-rate commissions are only good for a certain number of shares. You’ll be charged an additional amount (per-share) for all shares over that maximum number of shares. I may be wrong but I think it’s usually better to go per-share.

The other issue here is payment for order flow. That’s when a broker has a deal with a market maker to send their clients’ orders their way. That market maker can execute your order away from the best bid/ask. So commissions are not the only cost to take into consideration. I’ll gladly pay a slightly higher commission for super-quick executions and smart order routing. Sure you can find super low commissions but are you getting raped on the executions?

Sam wrote:

Your TA’s are real helpful and not to mention, you have a great site.
Question: Do you believe in having multiple brokerage accounts? For eg. TradingDirect offers $9.95 for up to 1,000,000 shares Would that not be helpful whenever one wants to trade large number of shares of a less expensive stock (say $5 or less)?

One can certainly come up with some good reasons to have multiple accounts:

• one for a certain trading style and a different account for another
• different brokers may have different short lists, so you may be able to short stocks at broker B that you couldn’t at broker A
• The ability to trade the same instrument on different time frames / directions. For example, you may want to short the S&P 500 for a scalp but be long it for a swing trade.

So if you have the money to spread around it may be a good idea. Personally, I like having everything in one place.