Williams Alligator: What the Three Lines Show

The Alligator is three moving averages of different lengths, displaced into the future. Its author named them jaw, teeth and lips, and named the indicator's states sleeping and hunting. Behind the metaphor is a simple idea: only trade when the lines have separated.

The Alligator indicator: how it is built

Three smoothed averages of the bar's median price: the jaw with a period of 13 shifted 8 bars forward, the teeth with a period of 8 shifted 5, and the lips with a period of 5 shifted 3. The forward shift is the key detail — the lines are drawn to the right of current price, like the Ichimoku cloud, which makes them leading reference points rather than lagging ones.

How often it signals on real data

We ran the rule "Alligator lines aligned upward in order" over daily bars for 12 currency pairs — 518 bars each, about two years. We counted frequency rather than profit: how often the state changes and what share of days the indicator spends in signal.

• State changes over the period: 43 (median across pairs)
• Annualised: 21.8
• Share of days in signal: 33.7%
• Sample: 12 pairs, 518 daily bars

The measurement is reproducible: the script and data are in the site's repository and the conditions are described in the methodology. Frequency is not return: it says only how often you will have to make a decision.

Sleeping and hunting

When the lines are intertwined the indicator is said to be sleeping — there is no trend. When they line up in order and diverge, the alligator hunts. The measurement puts a number on this: the lines are in ascending order on 34% of days. Given that roughly the same share belongs to the mirrored descending order, about a third of the time is left for intertwining — which matches what our ADX measurement showed.

What that gives you in practice

The Alligator is a state filter, not a source of entries. Its value is that it makes the absence of a trend visible, and that is the hardest state to recognise. Entering on the lines diverging is late by construction, because divergence is confirmed only after the move has started.

The forward shift and a testing trap

Displacing the lines to the right creates a concrete problem when testing on history. The value you see under today's bar was calculated from data eight days old — not from today's. If your tester pairs a bar's price with the line drawn beneath it, there is no error. But if you accidentally take the line that belongs to that bar by calculation, you are using data that did not exist at the time. The difference between the two versions of the code is one index, and it turns a losing system into a profitable one. It is the same look-ahead mechanism, hidden in the implementation rather than in the logic.

Limitations

Three averages mean three periods plus three shifts — six numbers, each of which can be tuned to history. The author's standard values work not because they are optimal but because they are fixed. Changing them turns the indicator into an exercise in curve fitting.

This material is for educational purposes and is not individual investment advice.