Fewer Trades, Better Monthseric

This is part thirteen of the ASX order book series. The previous post covered the averaging-down death spiral, the source of most catastrophic losses. This post is about the source of most slow, grinding ones: too many trades.

When I lined up the monthly dashboards from the journals I read, one pattern was so consistent it became a running joke in my notes. The best months had the fewest trades. Not slightly fewer. Often dramatically fewer than the trader's average. And the worst stretches were frequently the busiest.

Why more trades usually means worse results

If your setups vary in quality (and they do), each extra trade you take is, on average, lower quality than the ones you already took. The best opportunities of the day are usually obvious. The tenth trade of the day is usually not.

Here's a synthetic illustration. Rank a trader's daily opportunities by quality and look at the expectancy of each one.

The cumulative curve peaks at around three or four trades, then falls. A trader taking ten trades a day earns a quarter of what they'd earn by stopping at four, while paying two and a half times the brokerage and spending far more attention.

Real data is noisier, but the shape shows up in almost everyone's journal once they look.

The other costs of trading more

Beyond expectancy, extra trades carry costs that don't appear in the per-trade numbers:

• Commissions and spread on every one, as covered in the commission drag post.
• Attention. A trader managing a marginal position misses the A+ setup that appears while they're busy.
• Decision fatigue. Later decisions in a long session are worse.
• Emotional carry-over. A string of small losses in marginal trades sets up tilt and revenge trading.

Trade caps: the blunt tool that works

Some of the most dramatic improvements I saw came from a single rule: a hard cap on the number of trades per day, set at a small number. The pattern afterwards is typical: a much lower trade count, a similar or better win rate and a far higher payoff, because the cap forces you to wait for setups you really like and to manage those properly. Tracking your adherence, even as a simple count of consecutive days within the cap, makes it stick.

Caps work because they convert a vague intention ("be more selective") into a constraint that has a cost. If you only have a few bullets, you think hard about each one.

Variants that showed up across the journals:

Time of day matters too

A second theme that ran through every journal: most of the edge was in the morning. The open and the first hour or so carried the information, the volume and the clear setups. Midday was choppy and range-bound. Several traders repeatedly noted that their afternoon trades lost money, and some formalised it as a rule to stop trading after a certain time unless something exceptional happened.

That's consistent with how the ASX trades: the opening auction concentrates overnight information, and the closing auction concentrates end-of-day flows. In between, liquidity thins and moves are noisier. More experienced traders sometimes had different playbooks for different parts of the day, but almost nobody traded the same way all day.

"No trade is a win"

One mindset shift appeared in several journals once the traders matured: celebrating days with zero trades when nothing set up. Sitting on your hands through a slow market isn't laziness. It's the cheapest risk management there is. In periods with few catalysts, a good trader's trade count should fall. If yours rises when the market slows, you're trading boredom.

Find your own limits from your data

You don't have to guess where your cap should be. If you journal your trades with timestamps, two simple analyses will tell you.

Copy import pandas as pd

def pnl_by_trade_number(trades: pd.DataFrame) -> pd.DataFrame:
    """trades: columns [date, entry_time, net_pnl]. Average P&L of the 1st, 2nd, ... trade each day."""
    t = trades.sort_values(["date", "entry_time"]).copy()
    t["n"] = t.groupby("date").cumcount() + 1
    return t.groupby("n")["net_pnl"].agg(["count", "mean", "sum"])

def pnl_by_hour(trades: pd.DataFrame) -> pd.DataFrame:
    t = trades.copy()
    t["hour"] = pd.to_datetime(t["entry_time"]).dt.hour
    return t.groupby("hour")["net_pnl"].agg(["count", "mean", "sum"])

If your fifth and later trades have negative average P&L, your cap is four. If your trades after a certain hour lose money, your window ends then. It's rarely more complicated than that, and the answer is almost always more restrictive than you'd like.

When to relax the cap

Caps are training wheels as much as rules. Some of the more experienced traders traded more on days with heavy news flow and many genuine catalysts, and less on quiet days. That's the ideal: your trade count tracking the opportunity set, not your mood. But you earn that flexibility by first proving you can stick to a hard limit.