Understanding 'Risk Management & Trading Psychology'

Lesson -> The Position sizing for Active Trader (Part 03)

13.1 - Choose your own path

In the previous chapter, we addressed a crucial concept. We examined three different models to determine how equity can be determined. Each model has its own position sizing rules, but that is not enough. A standalone method is needed to determine the size of a position. We will now discuss Van Tharp’s methods for position sizing.

Three core position sizing methods are what I would like to discuss at the moment. They are:

  1. One unit per fixed amount
  2. Percentage margin
  3. Volatility percentage

These models can be used independently of assets and are not dependent on time frames. What does this mean? This means you can use these position sizing methods to any asset. This could include stocks, currency futures, commodity futures or stock futures. You can also use them in any trading session or trades that last for more than a few months.

This will help you understand the concept better. You could choose a simple trading system such as a moving average crossover or a basic trading system. You will need to identify entry and exit rules, and then calculate the return you would generate for that time period. You can apply one of the position-sizing methods (which we'll soon discuss) to the data, and then you can evaluate the performance. You'll notice a significant increase in stability and P&L.

This will give you an idea of the complexity.

  • Let's say you have a trading strategy - a simple cross over moving average system
  • This will allow you to start trading any signal generated by the system.
  • There are three models that define equity, and at least three models that define the position sizing methods.
  • This means that you can position size in three x three = nine different ways to use cash for the same opportunity (signal).
  • Each will have a different P&L.

My experience has shown me that one method is best for estimating equity, and maybe 1 or 2 (meaningful) techniques to determine the position size. It is possible to be more complex than you need, but it does not always mean that the result will be better.

As a trader, you need to determine which path to take based on your personality. Let's move on to the core position sizing methods.

13.2 - One unit per fixed amount

Let's first discuss the Unit per Fixed Amount model. This model is quite simple. Anyone who is interested in position sizing and trades would have looked at this model. This model is simple and I love it as well as dislike it.

You simply need to state how many shares (or lots, in the case of futures), you plan on trading for a certain amount. Let's take, for example, Rs.2 lakhs in your trading account and the following 5 assets (futures), which you use as your opportunity universe.

  1. Nifty
  2. SBI
  3. HDFC
  4. Tata Motors
  5. Infosys

It could be stated that you wouldn't want to trade more then 1 lot of futures for every 100,000 assets at any one time. Assume you receive a signal to purchase Nifty. Now, there are 2L in your account. You can choose to purchase one or two lots.

This model does not make it difficult to make decisions. This model has few flaws.

This is what you should think about: The trading system you follow signals you to purchase Nifty Futures, while the system signals that you need to buy Tata Motors. You have 2L in your account so you decide to purchase 1 lot each. Note that Nifty Futures needs a margin of around 60K, and Tata Motors about 72K at the time of writing.

The rule states that 1 lot is equal to 1 L, regardless of margin. The position sizing rule means that both contracts are given equal weight, disregarding the implied riskiness of the asset. For example, Nifty Futures' annualized volatility is around 14% while Tata Motors' annualized volatility is over 40%. You are thus exposing yourself to higher levels of risk at the portfolio level.

This is actually both good and bad. It rejects trades based on riskiness, but it is bad because it doesn't really factor in risk.

Another angle is available. Consider this: If you follow a trading system, you will apply the 1 lot per 100,000 size rule. Let's say you have 2 lac capital. Assume that the system works well and you receive multiple winning trades. The maximum number of lots that you can purchase for each signal is limited to 2. To increase your lot size, you will need to either double your capital or wait until your profits double. This particular method of sizing a position limits the system's scalability. This can be overcome by bringing in a larger account.

These are the reasons why I don't like the 'unit per set amount' position sizing method. But don't believe me? I suggest that you do some research and find out what your comfort level is with this technique before making a decision to adopt it or not.

13.3 - Percentage Margin

The percentage margin is a great position sizing tool. This technique, especially for intraday traders, I believe is more structured than the "unit per fixed amount" technique. You must position size based upon the margins using the percentage margin technique.

This is basically how you fix a certain percentage of your capital to be used as a margin amount for a trade. Let me show you a illustration to better understand the concept.

Let's say you have Rs.500,000/- capital. You decide to not expose more that 20% of your margin amount to any trade. This would translate to Rs.100,000.

If you see an opportunity to trade Nifty Futures you will be able to easily accept this position since the margins are approximately 60K. Let's suppose you see an opportunity in ICICI. You will have to give up this position as the margin is close to Rs.105,000/. ICICI will therefore be excluded from your trading universe unless and until you increase your capital. You should not increase capital randomly to take advantage of opportunities. Profits accruing in your account should result in capital increasing.

Suppose you spot an opportunity in ACC after you have initiated the Nifty position. The margin for this is 90K.

Do you agree to this?

This really depends on how you value equity.

You can take the position in ACC if you look at the total equity model.

However, this model would be more efficient if the total equity is reduced (assuming a 20% position sizing rule).

Start Capital = 5L

Margin block = 60K

Capital new = 4.4L

Margin @ 20% = $88K

This means that you would be short (just) 2K to fill a 90K job. You would then have to give up. Equity estimation is a crucial part of this process.

Let's say you see an opportunity that requires a margin greater than 40K. Since you have 88K, it is possible to comfortably take up two lots.

And so on.

You pay approximately the same percentage to all positions as you do for other positions. Each position's volatility could be different. This could lead to you placing risky bets that can alter the overall risk profile of your account.

Next position sizing models reduce this risk.

13.4 - Percentage Volatility

The percentage volatility rule is used to account for volatility in the underlying asset. This technique uses the daily expected movement of the underlying to calculate volatility. It is not the "standard deviation".

If SBI's OHLLC is 276, 274, 274, 274, or 278, then volatility for the day simply refers to the difference between low- and high, i.e.

279 - 274.

= 5

This gives me an idea of generic volatility. I can simply look at the difference between high and low for the last 'n days and then take an average. The problem with this approach is that it would ignore the gap up- and down openings. Van Tharp suggests that you use the 'Average Tru Range' to measure volatility in stocks.

Position sizing using the 'Percentage Volatility’ method requires us to determine the maximum volatility exposure for the equity capital.

If the equity capital is Rs.5 lakhs let's say I don't want more capital exposed to volatility than 2.2%.

Let's look at an example. Here's the Piramal Enterprises Limited chart (PEL).

Image 01

The 14-day average daily turnover ratio is 76. This means that each share of PEL contributes a fluctuation (volatility of Rs.76/) to my equity capital.

Let's say I spot an opportunity for trading PEL. Now, I need to decide how many shares I should buy. My equity is 5L. I have capped volatility exposure at 2%.

2% of 5L equals 10,000/-. This means I should limit my holdings of PEL to 10k shares.

This means that 10,000 can be divided by 76 to get the number of shares I can purchase.


= 131.57, or approximately 131 shares.

PEL trades at 2700. This means that your total exposure would be -

131 * 2700


For estimating equity, I recommend that you use the reduced total equity method. This would mean that the capital available to trade the next trade would be:

500,000 to 353,700


At 2% volatility, capital exposure has been reduced to Rs.2929/+. The capital exposure to the next trade will decrease, but volatility exposure would not change.

Van Tharp has some advice. If you're inclined to use percentage volatility, you should also estimate how much volatility you wish to expose your portfolio to. On a 5L capital, this would be Rs.75,000/-.

Consider this: If every position is against you, you could lose 75k on capital of 5L in a single day. How does that make you feel? 15% portfolio volatility may be too high for you if your stomach hurts.

We will be exploring a few more concepts in the next chapter before moving on to understanding trading biases.

To Summarize

  1. Position sizing is a process that involves estimating equity.
  2. Let's say you have three ways to estimate equity, and four ways to size your position. This will give you a total of 4 x 3 = 12 positions sizing method.
  3. You will need to determine how many shares you will trade for each 'x' amount capital in your account using the unit fixed model
  4. Unit fixed model does not consider risk
  5. The percentage margin method will require you to determine the maximum amount of capital you will expose to. This should be combined with total reduced equity model
  6. Percentage volatility is a measure of volatility in terms or ATR.
  7. Weights of 'volatility exposure' to each position equal percentage volatility