Microcap execution costs - what annual drag do you actually see?

[tgoudie000]

For those running microcap strategies, I’m trying to understand what you see in real execution for total implementation cost and how you model it in P123/backtests.

For estimation sake assume roughly a $1M North American portfolio trading long and short microcaps, e.g., MedianDailyTot(126) > 50000, Price> 1, and MktCap > 10, with say 50-100 positions.

I’m interested in your estimated annual cost for each of these separately as %/yr:

• Entry/exit price: Open vs Close vs HLC2/4 or another execution-price assumption. (I use HLC2/4 Monday)
• Bid/ask spread (I use advanced 0.25)
• Market impact (I use advanced 1.0)
• Carry: short borrow fees and/or financing costs (I use 3%)
• Commissions/other transaction costs (I use 0)

For example, if my theoretical long/short HLC2/4 strategy returns a CAGR of 50-75% before execution costs but Trades on average every two weeks and rebalances weekly, my CAGR drops to about 5% after I account for execution costs; how many percentage points per year do you realistically see for slippage and impact in live trading, and does that drive a holding period of at least 3- 6 months to deal with the slippage and impact in microcaps without destroying CAGR?

I’m especially interested in actual experience trading microcaps, rather than generic cost estimates, as I am modeling using HLC2/4 & advanced Slippage of K=1 and I=0.25, and I have NO idea if that is reasonable for the universe I listed above.

If you were modeling a $1M long/short microcap portfolio today, what total annual execution drag would you assume for Slippage and Impact, and how does that affect your execution and trading strategies?

[yuvaltaylor]

@tgoudie000

For those running microcap strategies, I’m trying to understand what you see in real execution for total implementation cost and how you model it in P123/backtests.

For estimation sake assume roughly a $1M North American portfolio trading long and short microcaps, e.g., MedianDailyTot(126) > 50000, Price> 1, and MktCap > 10, with say 50-100 positions.

I’m interested in your estimated annual cost for each of these separately as %/yr:

• Entry/exit price: Open vs Close vs HLC2/4 or another execution-price assumption. (I use HLC2/4 Monday)
• Bid/ask spread (I use advanced 0.25)
• Market impact (I use advanced 1.0)
• Carry: short borrow fees and/or financing costs (I use 3%)
• Commissions/other transaction costs (I use 0)

For modeling in a Portfolio123 backtest, I would suggest using advanced slippage with the following parameters:
Spread (L) = 0.2
Impact (K) = 0.35
Impact Basis = Starting Capital
Starting capital = $1M if that's what you're using.

Also, add this to your universe rules: SlippageAdv(1000000) < 100.

These numbers are based on my actual slippage costs, which I've assiduously measured over the years.

If you're not using VWAP orders, raise L to 0.25 and K to 0.45.

[Jrinne]

@yuvaltaylor

If you're not using VWAP orders, raise L to 0.25 and K to 0.45.

Do you use the Average of High, Low and 2X Close or the Open in your sims with these L and K?

Thanks!

[yuvaltaylor]

@Jrinne

Do you use the Average of High, Low and 2X Close or the Open in your sims with these L and K?

Thanks!

Average. I never use Open.

[Whycliffes]

@yuvaltaylor

If you're not using VWAP orders, raise L to 0.25 and K to 0.45.

I am a little uncertain, and I may have misunderstood something here, but why do you believe there is a greater impact (K) when using VWAP orders? Isn't the purpose of this type of order to result in a totally smaller market impact?

[yuvaltaylor]

@Whycliffes

I am a little uncertain, and I may have misunderstood something here, but why do you believe there is a greater impact (K) when using VWAP orders? Isn't the purpose of this type of order to result in a totally smaller market impact?

K = 0.35 if you use VWAP and K = 0.45 if you don't.