(Adapted from work by Kurtis Hemmerling)
PDF Version: Choosing Portfolio123 Designer Models Supplement – Liquidity Considerations
For many reasons at various times, Designer Models subscribers may have reward-risk preferences that lead them to aim at smaller capitalization issues having less trading liquidity. Whether this is a good or bad thing to do depends entirely on ones understanding of liquidity risk and one’s effectiveness in managing it. This Supplement aims to help in both respects.
Allocating Capital Per Model
Much of the liquidity profile of your portfolio will result from the decisions you make regarding how much capital you choose to allocate to each of your models?
From the vantage point of trading liquidity, this issue tends to be benign with large-cap and often mid-cap stocks (investors there are free to focus on ordinary financial return-risk matters). Liquidity risk becomes most pronounced in the small- and micro-cap areas of the market. For those who choose to focus here, an important question is raised: What is the maximum amount of capital that an investor could put in a model without extremely impacting the prices negatively?
A basic rule of thumb, particularly for those less experienced in trading, would be 5% of the daily turnover.
The daily turnover is the share price multiplied by the volume. So if a stock has a turnover of $100,000 per day, you could buy or sell $5,000 without too much trouble if you used sound practices such as limit orders. So, repeating our Rule of Thumb: Trade up to 5% of the daily turnover.
How can you quickly discover the daily turnover in each model?
Portfolio123 has provided the daily turnover of the lowest liquidity stocks in the portfolio – the bottom 20% of the holdings. Why would Portfolio123 only give the average turnover statistics of the bottom 20% instead of the entire portfolio? Imagine you had a model with huge liquid stocks such as Google and Apple mixed in with some thinly traded microcaps. Your average turnover amount would not be meaningful when trying to buy and sell the thinly traded stocks and would give a false sense of confidence.
You can see this as part of the Designer Model page section covering “Trading Commitment”. It’s located among the Statistics presented on the right side of the page below information about the Designer. pages.
Here is one example of a theTrading Stats where the Buy Daily Average for the bottom 20% of the portfolio is $273,234 per day. Five percent of that number is $13,661.70 – which would be the maximum suggested amount of capital per stock for a new investor. If this portfolio has 7 stocks, the amount of capital to invest could be up to $95,631.90.
Before you run out and set your maximum portfolio size based on the average daily turnover, there are a few other considerations that you need to keep in mind. Your ability to efficiently buy and sell stocks without undue slippage will also be influenced by the depth of liquidity and the average amount of days a stock is held.
Liquidity Depth
Another basic rule is that smaller capitalization stocks often have a shallower depth of liquidity. What is liquidity depth? Consider how two stocks trade.
- Stock ABC is a microcap stock with average daily turnover of $100,000. The bid is at $10 and the ask is at $10.05. You want 500 shares. You buy 200 shares at $10.05, 100 at $10.20, 100 at $10.50 and the final 100 shares at $10.75. Your average share price was $10.31 meaning your slippage was a whopping 3.1%
- Stock XZY is a large cap stock which also has an average daily turnover of $100,000. The bid is at $10 and the ask is at $10.05. You want to buy 500 shares. After buying the 300 shares at $10.05 more shares become available at $10.10 and $10.15. Your average price is $10.08 or 0.8% slippage.
While both stocks have the same average daily turnover and the same bid/ask spread, stock XYZ has more liquidity depth. When buying shares at the ask in stock XYZ, more shares become available at a nearby tier. It may very well be that you can get a better average price(less slippage) in a large cap stock than a smaller cap stock with the same liquidity statistics.
Average Days Held
Another factor to consider is how many days on average a stock is held. How will this affect your trading?
- If the stock is held an average of 7 days, you will need to quickly buy the shares on the Monday it is recommended.
- If the average holding time is 3 months you easily build a position over many days. Thus, you may be able to buy up 10 or 15% of the average daily total when spreading your order out.
So to modify the rule of thumb, you can buy up to 5% of the average daily turnover each day that you accumulate a position . . . if you are careful. But you should also be aware that this makes unwinding a position quickly more challenging. You may need to sell over a period of days as well.
Trading the Portfolio
You have determined the maximum amount of capital you can trade in each model – or at least you have a good idea. What are some good practices to follow when actually buying and selling stocks recommended by a model?
Determining Model Slippage
Before you set out to buy all the positions in a portfolio, you should be aware of the parameters used by the model designer. How much slippage did they factor into the model when backtesting their strategy? What is the maximum amount of slippage you can accept in order to achieve similar results to the Designer Model?
Traditionally, model designers had two options when factoring the cost of slippage in their models.
- They can add a fixed slippage amount (e.g. 0.15%)
- They can select a variable slippage which is adjusted internally by Portfolio123
Variable Slippage
Below is a guideline for the variable slippage formula used by Portfolio123.
- If the bottom 20% of stocks in a portfolio have an Average Daily Turnover of $200,000, the variable slippage would be 0.75%.
- If the bottom 20% had a turnover of more than $5 million the variable slippage would be 0.10%.
- In addition to this variable slippage amount, Portfolio123 has added one penny to the total. Why? Imagine you are trading a super liquid stock where the bid is $1 and the ask is $1.01. Even though variable slippage would suggest 0.10%, you would lose a minimum of 1% if you bought at the ask…which is one penny. So in order to more accurately reflect the slippage of lower priced stocks, one penny was added.
Nowadays, model designers can continue to use any slippage assumption they wish for their personal models, but they are required to use Portfolio123 variable slippage for any model that is submitted to the Designer Platforms model.
Alternative Brokerage and Cost Per Trade
Calculating brokerage fees is not feasible for a publicly traded model because these fees can vary widely and Portfolio123 cannot know what arrangements an individual subscriber mat have (unless that subscriber chooses to use Interactive Brokers, a firm that interfaces directly with Portfolio123). So you will need to mentally factor this in yourself.
Acti-quant 