Evaluating percentage profit exits on trend-following averages

A second close on the same trend rule

The archive kept a single trend-following procedure and then added profit-taking as a second close. A long or short was taken on the close when an exponential moving-average of the close turned. The hold-until-the-average-reverses path stayed in place as the control.

Profit-taking closed the open trade when unrealized gain on the close exceeded a preset percentage of that close. The two closes belong to one procedure: the average still supplies the directional state, and the percentage target only decides whether the active position is taken off before that state reverses.

One moving-average, several speeds

The moving-average converts an ordered close series over a stated lookback into a directional state, so the same trend rule can be compared with alternative exits on the same sample. Lookbacks of 5, 10, 15, 25, 50 and 75 days were tested.

The smoothing-constant was set as two divided by lookback days plus one. Each lookback therefore defines a distinct trend speed, and that speed is the first axis of the evaluation grid.

Entry, profit-taking and same-direction-lockout

A long or short was taken on the close when that average turned. The trade was closed on the close if unrealized profit exceeded a chosen percentage of that day's close.

After a profit-exit, the procedure did not open another trade in the same direction. Implementations kept a trend-direction column separate from an active-position flag. Same-direction-lockout keeps the trend state visible and leaves the position flat until the average itself turns the other way.

A halved target grid and a no-take-profit control

Targets of 100%, 50%, 25%, 12.5%, 6%, 3% and 1.5% were compared. The 100% level was used as a no-take-profit control because no profits were taken there. That control is the hold-to-turn path. The smaller levels are the profit-taking overlays on the same moving-average rule.

Profits scored against a stated risk unit

Results were scored as profits versus risk. Risk was defined as one standard deviation of daily accumulated profits and losses after commissions and slippage. That score is the risk-reward-ratio used to keep a loss or exposure reading bounded while each lookback and each target is compared.

The same grid on three samples

The same lookback-by-target grid was tabulated for Hang Seng, Deutschemark and Unilever samples spanning late-1980s through early-1990s windows. The Hang Seng reward-to-risk sheet is the map the evaluation text asks the reader to learn to read.

How the Hang Seng map lines up

On the Hang Seng reward-to-risk grid, larger percentage targets lined up with slower lookbacks and smaller targets with faster ones. The pairings were 12.5% with 75 days, 6% with 50 and 25 days, 3% with 15 days, and 1.5% still favored for the fastest speeds.

Editorial reading: use that map to set target size with trend speed. A slower average is paired with a larger percentage. A faster average is paired with a smaller one. The sheet is a design map, not a prompt to wait for a rare runaway move.

What the evaluation text tied to lower measured risk

The evaluation text tied lower measured risk to less time in the market, exits taken while price was still moving in the trade's favor, and avoidance of late-move volatility rather than to holding until a reverse signal.

A small realized move inside a wide range

In one currency-chart illustration, marked long entry and exit points captured 200 points from an 800-point range. The illustration showed that high later volatility can demand a large stake for a small realized move.

Editorial reading: that picture is why a late-move hold is not treated as the default design. The hold-to-turn control stays in the grid so the smaller targets can be judged against it, not so the large later range becomes the habit.