Don't Confuse a Bull Market With Skill

Buy almost anything in a rising market, hold it a while, and it pays. It pays when the reason for buying was poor, when there was no stop and no plan for putting one in, and when the size was wrong, because the market covers all of that, and the trader responsible concludes, reasonably enough, that the method works.

The conclusion may well be correct. It is also untested, which is a different thing entirely.

The pattern is familiar enough from the outside. A trader has a year in which, by their own assessment, they are very good at this. The index was up rather a lot that same year.

The people around them, who have never held a position in their lives, tend to remember the period chiefly as the one they would not stop talking about, and are usually some months ahead in reading the situation correctly. That gap, between the account statement and the view from across the table, is the subject of this post.

Tide and swimmer

In a bull year, most of the return is the market’s. Whatever your entries and exits added or subtracted is a smaller number sitting on top of a much larger one, and the larger one belongs to everybody.

There is a quick way to see the split. Take the trading return for the year, then subtract what an index fund would have made over the same period at the same average exposure, a calculation that takes an evening and gets put off for years. What survives that subtraction is the part your decisions contributed. For a good many traders in a good year the figure is modest. For a fair number it comes out negative. They made money and still did worse than sitting still would have done.

What each regime reveals

A trending bull market rewards holding, punishes tight stops, forgives poor entries, and conceals sizing errors for the simple reason that the drawdowns which would expose them never arrive.

A range-bound market does something else. It rewards fading the extremes, punishes breakout buying, and finds out anybody whose method has no rule for the absence of a trend, which is a rule most methods never troubled to write down. This is where trend followers hand back a bull year’s gains, one small loss at a time.

Then a bear market. It punishes everything held long, exposes correlation you had no idea you were carrying, and settles the question of whether the stops and the sizing were real or decorative. Traders who come through one tend to emerge with a lower opinion of their own skill and a considerably better risk process, and the two changes generally arrive together.

A strategy stays untested until it has run in a market that does it no favours. For a long-biased method that means two markets. A sideways year, where entries are stopped out by chop. A falling one, where the tide pushes against every position you hold.

An illustration

Picture a first year of trading that lands in a strong market. The results are excellent. The trader is confident the results are theirs. The year after is sideways, the method is almost entirely trend continuation, and what it produces is a long procession of small stop-outs, not one of them large enough to alarm anybody until they are added up at the end of the quarter. The solid-looking expectancy stays negative for the better part of eight months.

The arithmetic afterwards is uncomfortable. That first year’s return, less the index return over the same stretch and adjusted for how much was actually invested, comes to a modest positive number, a great deal smaller than it had felt at the time. The second year, in a market that offered it nothing, shows what the method was worth alone.

Finding out what is yours

1. Benchmark every year against the index over the same period, scaled to average exposure. The gap is your contribution. The rest is tide.
2. Backtest across regimes deliberately. Pull out a sideways stretch and a falling one and run the rules over those before trusting anything a rising market has told you, since a rising market will tell you more or less whatever you would like to hear. The backtesting guide treats regime dependence as one of the five biases.
3. Track win rate and expectancy by market condition in the journal. If everything good happened while the index sat above its 200-day average, what you hold may be a filter rather than an edge, and knowing which one it is matters a great deal.
4. Write the no-trend rule now, while the market is kind.

None of which makes a bull-market return imaginary. The money is real. What wants discounting is the confidence it buys, along with the size that confidence encourages once the weather turns, which it does without consulting anybody.