Written and copyright © 2009-2013 by Thomas N. Bulkowski. All rights reserved.
A long time ago, I decided to test whether or not scaling out of a trade made sense. Scaling out means selling not thy whole wad at one time, but selling only a portion of your holdings
each time. For my trades, I determined that you lose more money than if you just sell the entire position at once.
This article takes a closer look at scaling out using two methods, one based on sample trades and another based on an article I read.
Scaling Out: Summary
If you want to make less profit, then scale out of a trade. If you want to make more or retain more of your capital, then do not scale out of a trade. Just sell the entire
position at once.
Scaling Out Opening Position: Method 1
The September 2009 issue of Active Trader magazine has an article titled, "Scaling out as an exit technique," by Howard Bandy. I will get to his conclusions in a moment.
But first, I programmed an Excel spreadsheet just to crunch the numbers on a series of hypothetical trades. They appear in the table below.
| ||Test 1||Test 2||Test 3||Test 4||Test 5|
|Scale out price|| ||$12.50|| ||$12.50|| ||$12.50|| ||$9.00|| ||$9.00|
|Best||BuyHold|| ||BuyHold|| || ||ScaleOut||BuyHold|| || ||ScaleOut|
In the first test, I bought 200 shares at 10 and sold it all at 15 (buy & hold), giving me a profit of $980 after deducting $20 for round trip ($10 for buying and $10 for selling) commissions.
Scaling out half of the position (100 shares) midway to the 15 target gives a profit of $720. Scaling out is worse.
A day trader I know that also teaches others to day trade recommends scaling out. On the first trade of the day, he sells a portion to lock in a profit, booking a win to place you in the
proper frame of mind. I can't argue with that from a psychological point of view, but I grab my wallet as I say that. It doesn't make much sense otherwise.
What you're doing is cutting your profits short instead of letting them accumulate. I think his advice is wrong, but that's just me. Perhaps novice traders need the mental crutch.
In the second test, price starts moving up from 10, climbs above 12.50 but does not reach the sell target at 15. A stop takes you out at 12.50 for a gain of $480 (buy & hold).
Scaling out half of the position at $12.50 costs you commissions, leaving you with a net gain of $470, so scaling out is worse.
The third test is similar to the last one. Price climbs from 10 to over 12.50 but does not reach 15. This time, though, your stop is at the entry price of 10 (break even), so you
lose the cost of commissions ($20)(buy & hold). Scaling out, however, means you sell half your holdings at 12.50 and the rest at 10 for a gain of $220. Scaling out
works better in this scenario.
Test 4 is when price drops. You buy 200 shares at 10 and it drops to 9, leaving you with a net loss of $220 (buy & hold). If you were to scale out at $9 on the hope that price would rebound
but it drops to 8 instead, that gives you a loss of $330, which is worse than just selling the entire position at one time. Scaling out is worse.
In the last test, you decide to hold onto a losing position longer, riding it down to 8 for a net loss of $420 (buy & hold). The scaling out method, selling half of the position at 9 saves
you money, booking a net loss of $330.
In three of five scenarios, scaling out costs you more money than selling your entire position.
In my trading experience, test 1 and test 4 are what happens. In the first test,
selling shares along the way up to a target costs you money. In test 4, hoping a losing position will see price rise again also costs you money when price continues lower. Selling the entire position
at once saves you from a bigger loss caused by scaling out.
You will want to evaluate which of the tests shown in the table applies to your trading.
Scaling Out: Method 2
Howard Bandy, in the Active Trader article mentioned above, conducted simulations to determine whether or not scaling out is beneficial. He used 72 stocks and ETFs chosen for their liquidity,
using the test period from December 31, 2003 through October 31, 2008, giving 4,176 trades.
His trading system consisted of buying long at the close of the first trading day of
the month and selling at the close of the last trading day of the month. Trading commissions and slippage were not factored into the simulation. He used a "scale increment" to determined
the price at which a sale occurred, and it was based on the standard deviation of closing prices over the prior 21 trading days, evaluated once before buying each security.
Then he tested the buy and hold method and the scale out method. The scale out method used various standard deviations ranging from 0.50 to 4.50 in 0.50 increments to set profit targets
used to scale out. He also used a maximum-loss stop during some of his tests. When it was used, the stop price was set at the price of the initial buy creating a break-even point. The
stop only took effect once a scale out trade occurred (meaning you could not be stopped out before the first sale). As a variation, he also tested placing a stop as soon as the buy occurred.
Scaling Out: Bandy Conclusions
This is what he concludes, for both winning and losing trades, "...the closer the profit target is to the entry price, the more the average profit is reduced." My interpretation
of this is that singles and doubles are fine,
but the big bucks come from the home runs.
"...Setting the maximum-loss stop to the break even price dramatically reduces profit." Once his system enters a trade in which it scales out,
he raises his stop loss to the initial entry price (break even). He concludes that using a break even stop forces the system to exit trades that would otherwise be profitable.
Looking at winning trades only gives the same conclusion. Scaling out of trades when the price target is near the entry price reduces profitability. Once you scale out of the trade,
using a stop loss set at break even hurts profitability.
What about losing positions? Scaling out reduces the size of the loss but it does not turn losing trades into winning ones.
Scaling Out: Maximum-Loss Stop
He then tests the maximum loss stop by using it as soon as the trade is entered. He finds that doing so "greatly reduces the profitability of the system." The farther away from the
entry price the stop is placed, the more profitable the system becomes.
This finding is not new. Profit drops because the stop takes you out of your big winners. I proved
that to myself when I looked at my own trades years ago. This is not an excuse to avoid using stops. Rather, it highlights the need to know when to sell and when to hold.
He also says that placing a stop 2 standard deviations below the entry price cuts profits in half! Wow. So if you use Bollinger bands each set 2 standard deviations away
from the moving average as an exit, you may be wondering why performance is so poor...
He concludes by saying that moving the stop to break even when scaling out reduces profits to one-third. Ouch.
Scaling Out: What Do I Use?
Do I scale out of a trade? Yes, but only to reduce the position size. For example, earlier this year (2009) I had over 70% of my portfolio in utility stocks. With the markets trending since
March, I decided to diversify and sold one utility stock completely and cut the remaining size of several others. That dropped the utility sector to 50%, which I consider still much
too large. However, I can't find anything worth buying recently, so I am just collecting dividends at a rate as high as 8%.
Other than that, no, I don't scale out. When I sell a position, I sell it all, especially if it's for a loss. Scaling out of a losing position is an excellent way to lose more money.
-- Thomas Bulkowski
Written and copyright © 2009-2013 by Thomas N. Bulkowski. All rights reserved.