Saturday, September 6, 2008 @ 5:03 AM

Thoughts on Position Sizing

From Market Barometers (Everyone) | Discuss | Print  

This topic was discussed in a Conversation with Charles Kirk on April 25, 2008:

Kirk:   One of the reasons (among many) that I asked you to come in for the Q&A was to talk about your techniques concerning risk management and the use of your proprietary InVivo indicators. One of the most confusing and frequently least understood, but at the same time the most important aspect of trading successfully, is proper risk management. However, if you ask 10 traders what they mean by risk management you’ll often get 10 different answers. Of course, many of their responses will frequently include “you must use stops,” to “keep your losses small,” to “never risk more than % amount on every trade,” etc. But, many traders and investors find these suggestions not helpful because they’re both incredibly difficult to consistently and effectively apply in the real world especially when real money and emotions are on the line. That’s why many traders have been flocking to systems and indicators that offer buy and sell signals mechanically, much like what you offer through your proprietary InVivo indicators. For those who are unfamiliar with these indicators, can you provide a basic introduction to them and give a general sense on how they work and how you use them?

Teresa Lo:   It seems to me that risk management has several components. First and foremost, how much leverage does the trader intend to use? I wrote a free book about trading called Own The Zone. In Chapter Four, I go over the logistics of trading and the use of leverage. If there is one single determining factor that can be listed as the “cause of failure” in trading, I think it has got to be the excessive use of leverage. For example, futures and forex traders routinely use 2-3% of their own money and borrow 97-98% of it on margin. This means that if they lose the 2-3% on a trade or a series of trades, they are out of the game. These traders cannot afford to swing trade off the daily chart because the normal fluctuation is too high, so they move into the intraday time frame where increased commission and slippage nickel and dime them to death. On top of all that, traders that are not aware of the leverage issue also tend to be novices that don’t know much about trading and/or they are not used to making a large number of decisions in a small amount of time that day-trading demands. It’s a recipe for disaster.

The second issue is position sizing. People always feel that this is an inherently complicated subject but if we talk about it in terms of a casino game such as roulette and how much of the bankroll to bet on each spin, I think most people instinctively know that it is unwise to bet it all on black and hope for a good outcome time after time. There has been a lot written about position sizing. The basic goal is to reduce the risk of ruin to some small probability. For example, if a trader uses no margin and loses 5% of his capital per trade, (ignoring the declining balance) it would take 20 losing trades for 100% of the money to be lost. Say a stock is trading at $30; the initial stop loss might be indicated at $28. A trader might allocate $5,000 from his account to trade each stock with a 5% risk budget. We make two calculations and trade the smaller number of shares:

Calculation #1

$5000/$30 = 166 shares x $2 loss = $332 minimum potential loss, which is larger than the risk budget.

Calculation #2
$5000 * .05 = $250 risk budget
$30 - $28 = $2 minimum potential loss
$250/$2 = 125 shares
125 * $30 = $3,750 to check that the position size does not cost more than the capital allotted.

I know other traders use Kelly, fixed fraction or even “R”, but I have serious reservations about these schemes and do not use them.

The third issue is diversification. A trader may have a $50,000 trading account that can be divided up into 10 or 20 positions, but he must ensure that the they are not all from the same sector. When we calculate the correlations, we don’t want the positions to be highly correlated to one market; otherwise, we might as well trade that market.

Recently, individual and professional investors have taken tremendous losses. When we take a close look, they probably mishandled all three issues and ended belly up.

Kirk:  Very good points on both the use of leverage and position sizing. I also know that you think stops “cannot be eyeballed” and that a “trader that places stops arbitrarily or uses stops calculated incorrectly is destined to always exit too early and miss the home run, or exit too late after a change of trend.” Can you provide us with three examples of how your indicators have provided an effective stop to trade?

Teresa Lo:   Let’s take a look at the most recent mania in gold through the streetTRACKS Gold Shares (GLD):

Gold

NOTE: Subsequent to publication, we introduced the Universal StopFactor as the default setting to make life easier.

I’ve applied InVivo.RMI stops (StopFactor 1.1) and histogram to the daily chart. A trader whose strategy is to trade with the trend would not go short so long as the price bars are green and the green histogram lines are above the grey threshold line. My indicator helped traders exit long positions after the uptrend ended in March.

The next chart is the PowerShares DB Agriculture Fund ETF (DBA):

Agriculture

I know the mass media is going crazy about food shortages, etc., but the chart of DBA is similar to GLD. Again, a trader whose strategy is to trade with the trend would not go short so long as the price bars are green and the green histogram lines are above the grey threshold line. My indicators helped the trader to exit long positions after the top in March.

The last chart is Google (GOOG).

Google

While investors were clinging to hope, the price action said otherwise. In December 2007, the green histogram lines went below the grey threshold line as the stock began to weaken. The sell signal came in mid-December and was in place for three months.

Kirk:   One of the common complaints about stops (especially in a volatile tape) is that you’ll often get stopped out on positions too early. How do these indicators address this problem?

Teresa Lo:   Well, it’s never going to be perfect, but I think it is wise to use stops, if only to keep a losing trade from slipping down the slippery “slope of hope.” A separate issue that plagues undercapitalized traders is the fact that they use stops that are way too tight because they “can’t afford” to place them in the proper location. If a stop is placed inside the normal fluctuation for the symbol and time frame, one must expect to be stopped out prematurely. The only answer to this problem is to place the stops where they ought to be and trade a smaller size.

Kirk:  I think you may agree that just like stops, entry points “cannot be eyeballed” and that you need to have a system in place that generates consistently good entries. It seems like the system you’ve created does both quite effectively.

Teresa Lo:  If we look at the GLD chart again, my indicators located entry points after each pullback while they also located an entry point after the January pullback in DBA.

Kirk:  Like most systems, everyone can certainly point out a few success stories here and there, but trading systems are only effective if they work with some relative consistency. In using these signals in your own work, what information can you provide as to how effective they’ve been in the past (either through personal results and/or backtesting)?

Teresa Lo:  Once the three components of risk management have been taken care of (leverage, position sizing, and diversification), we must then turn our attention to security and time frame selection. I wrote about it TWICE because people simply ignore this issue, but in the end, it all comes down to this. One important cautionary note when reading books or visiting system vendor websites is the presence of well-chosen examples. For some reason — and I blame infomercials — people have come to think that if they purchase a system, they should be able to point it at anything and cash will shoot out of the screen into their living room. We all know that this is not possible because even the quants on Wall Street are losing their shirts.

I am very picky about the stocks that I trade. They cannot be stock indexes. They have to be “in play”, meaning there must be enough buzz for me to find news stories that help me figure out what part of the sentiment cycle it is in right now.

Trading is not as easy as people believe it is. Stock picking is like trying to find a needle in the haystack: you had better own a really good metal detector. I know this sounds pessimistic, but after spending ten years online, I truly think that individual investors should view trading as a serious hobby with profit potential, but lower their expectations. Look at the performance of stock picker Bill Miller’s Legg Mason Value Trust ( LMVTX) of late. It goes to show that even fifteen years of beating the S&P might have been just a string of lucky coin flips. Or what about the CANSLIM Select Growth Fund (symbol CSSGX)? Why the disconnect between the theoretical performance and real life? Or Fidelity Magellan (symbol FMAGX)? Even Long Leaf is down 5% YTD.

Questions and Comments

Please LOG IN to join the discussion if you do not see the text box to enter your comments below.

 

Key Reference Articles

For your convenience, this is a list of key reference articles. Most of these articles are archived in the Knowledge Base for registered users and members.

Trading Ideas, Market Barometers, Correlation & Volatility

Investor Sentiment

Economic Fundamentals

Fundamentals of Trading

Investment Portfolios

Recommended Reading

Before you trade or invest, take time to contemplate and understand some fundamental truths.