Gold, 15 April 2013: the day the trend didn't wait

On Friday 12 April 2013, gold broke below a level around $1,520 that had held for the best part of two years. On Monday 15 April it fell roughly nine percent in a single session — from about $1,480 to close near $1,350 — the biggest one-day drop since the early 1980s. Two days took the price from the low $1,500s to the mid $1,300s. Nobody who was long got a pullback to sell into.

It is worth replaying for one reason: it is the cleanest example on the chart of a market that does not give you a second chance.

Setting it up

Start the session on Wednesday 10 April 2013 on XAU/USD, so you have two ordinary days before anything happens. Use tick-by-tick printing; on a daily chart this is two red bars and it teaches nothing. On the one-minute build, Friday’s break and Monday’s collapse arrive the way they arrived — as a series of small, survivable-looking moves that keep not stopping.

Open a second chart on the 1-hour timeframe so you can watch the higher timeframe candle refuse to close while the lower one keeps printing. Both are on the same clock, so neither can show you anything the other could not.

What it looks like from the inside

Friday’s break is the interesting session. The level had been defended repeatedly, and the first move through it looks like every previous false break — the kind of dip a buyer had been rewarded for buying all year. The difference is only visible afterwards.

Monday opens lower and does not bounce. Every rally that looks like the start of a recovery lasts a handful of candles. If you were waiting for a better exit, the better exit never came; the only good exit was the first one.

What to test

Three things, each in its own session with the same strategy name so the Analytics page groups them:

1. A fixed stop, honoured. Place a long on Thursday with a stop below the level and do nothing else. Let the tool close it. Notice how ordinary the candle that hits it looks.

2. The “wait for a pullback” exit. Same trade, but you decide when to get out. Record what the journal says about your MAE — how far the trade went against you before you acted.

3. Shorting the break. Enter on the Friday break with a stop above the level. This is the trade that worked, and it is also the trade that would have failed on every previous test of that level for two years. The point is not that shorting breaks works; it is that you cannot tell the difference at the time.

The honest caveat

You know this one ends badly for longs, which makes the short look obvious in a way it was not. The value is not in the result. It is in watching, minute by minute, how a market that is about to fall nine percent looks exactly like a market that is about to bounce — and in seeing what your hands do about it.