6 / 6 · 9 min
Trading waves without fooling yourself
Of the five waves, two and a half are tradeable. Everything else rests on guidelines, and a guideline gives you no price at which you admit you were wrong.
The course is nearly over, so it is time to assemble it into an order of operations. The main filtering has already been done for us by the three rules: it only makes sense to trade where a rule gives an INVALIDATION PRICE. Not an expectation, not a target, not a usually-it-goes-like-this — a specific number at which the count is declared wrong and the position is closed without discussion. There are exactly two such places.
Where an invalidation price exists and where it does not
| What is traded | Where invalidation comes from | Usable |
|---|---|---|
| entry after wave two, playing for wave three | the start of wave one — a rule | yes |
| entry after wave four, playing for wave five | wave one's territory — a rule | yes |
| entry in the middle of wave three | no rule says where it ends | no |
| trading the correction itself | corrections have no hard rules at all | no |
| catching the reversal after wave five | guidelines only | no |
Why the entry after wave two is the main one
It is the only one where the invalidation price sits CLOSE to the entry. By rule, wave two does not pass the start of wave one; so entering right after it completes gives you a stop a few percent away rather than twenty. This is not about the third wave being the strongest — it is about this being the cheapest place to be wrong. The entry after wave four is legitimate too, but invalidation is further away and the ratio is worse, which is why it is second rather than first.
A whole trade, from count to size
Wave one ran from 1,720 to 1,905, wave two pulled back to 1,762 and completed as a three on the lower timeframe. Entry at 1,768. The invalidation price is the start of wave one, 1,720; the stop goes slightly below, at 1,714, so it is not taken out exactly at the level. The distance is 54 points, about 3% of price. The target is calculated not from Fibonacci but from a rule: wave three cannot be the shortest, so it is at least as long as wave one — 185 points from the end of wave two, that is 1,947. The result: risk 54, move 179, a little over 3R. The trade qualifies not because it is pretty but because all three numbers are known BEFORE the entry.
On a $5000 account risking 1% per trade that is $50. Divide by 54 points and the position comes out at roughly $1637. Notice what is NOT in this: not a word about how confident you are in the count. Size is computed from the distance to the stop, not from confidence. Confidence is a feeling, and it grows precisely when the count is convenient — that is, when it has been fitted.
The order of operations on every trade
- 1identify the higher degree: where the large move is going — only so as not to trade against it
- 2label the working degree and find a completed wave two or wave four
- 3check the three rules by numbers; if even one is broken there is no count, move on
- 4write down the invalidation price and the wave number BEFORE entering, as a separate line
- 5compute position size from the distance to the stop, not from confidence
- 6after closing, record the outcome and, separately, whether you changed the count mid-trade
The value of this order is how often it says do not enter
Walk through the six steps honestly and most charts drop out: rules broken, wave two not finished, higher degree against you. That is the method working. Wave analysis adds no edge by itself — it adds the discipline of refusal, and that part is measurable: count how many of a hundred counts reached an entry. If more than twenty did, you are playing along with yourself.
Placing the stop by the target instead of by invalidation
A trader projects wave three to 1,947, sees a good ratio and moves the stop closer to make the ratio even better. The ratio improves on paper and gets worse in life: the stop now sits where the count is still ALIVE, and an ordinary swing will take it out. The invalidation price is set by a rule and does not move to make an arithmetic look nicer. The only thing you may move is position size.
Entering wave three because it is the strongest
Wave three is usually the longest is a guideline, not a rule, and more to the point it says nothing about where wave three ends. An entry in the middle of wave three is an entry with no invalidation price: the nearest rule (wave four does not enter wave one's territory) gives a number far below, and a stop there equals half the move. On paper such a trade looks the best of all, because the target is taken from a guideline and the risk from hope.
The only way to learn whether your wave count works is to keep it in a trade journal together with the invalidation price, and then look at the outcome across all entries rather than the memorable ones. That is what a practice account is for: it computes the result for you and does not let you edit the entry in hindsight.
Practice account and journalTwenty counts in replay with a written invalidation price
Work through twenty episodes in bar replay. For each, before moving on, write four things: wave number, degree, invalidation price and intended position size. Then roll forward and record the outcome in R. At the end compute three numbers: the share of counts that reached an entry; the average R across the ones you entered; and how many times you caught yourself wanting to relabel. The third number matters most: it predicts your result on a real account better than the first two.
Why is an entry in the middle of wave three worse than one after wave two, given that wave three really is often the strongest?
Because the point is not the strength of the move but the distance to the invalidation price. After wave two invalidation is near — it is the start of wave one, and being wrong costs a few percent. In the middle of wave three the nearest rule gives a number far below, so an honest stop is enormous; and a small stop would sit where the count is still alive and get taken out by noise. The strength of a move does not compensate for the absence of a place where you admit you were wrong.
In short, what is measurable in wave analysis and what is not?
Measurable: the three rules and the invalidation price they give; the distribution of retracement depths on your instrument; the share of counts that reached an entry; your own tally of relabelling. Not measurable: whether a count is correct in the moment, the strength of wave three, the crowd psychology usually invoked to explain waves. The first list is what a trade can be built on. The second is what makes conversations about waves endless, and for that reason it must stay out of the decision.