Forex Trading Strategies: Eight Styles and What Each One Costs You

By Uladzimir Leuchankou, founder of Signal2Forex · 4 November 2026

IG client sentiment is the share of IG accounts that are long or short in a market — EUR/USD, gold, the Nasdaq, anything IG quotes. It comes as two numbers that add up to 100%. On 4 November 2026 the four markets most of my readers trade looked like this.

IG client sentiment on 4 November 2026, rounded to 1%, taken from IG’s public market pages. The classic reading is the opposite of the crowd — short EUR/USD, long USD/JPY. Below I explain why I would not touch two of these four rows.

Is 64% long a sell signal? Most articles about client sentiment say yes — anything above 60% in one direction means fade the crowd. I suppose that is an answer too. But for me a single percentage is not a signal. It is a count of people, and people stay wrong for weeks. What matters is which way the market is moving while they stay wrong.

Heads, not money

IG counts accounts, not money. One trader with 0.04 lots and one trader with 40 lots are one vote each. The result is rounded to the nearest 1% and refreshed through the day as positions open and close. You can see the headline number on IG’s public market page without an account. The split by hour, day, week and month sits inside the platform, on the deal ticket, and for that you need to log in.

The percentage is easy to turn into a ratio, and the ratio is what IG’s own guide used. 64% long means 1.8 long for every short. 66.6% means two to one. 75% means three to one. I prefer the ratio, because “two to one” sounds like what it is — a crowd — while “60%” sounds like a pass mark.

The old IG sentiment report is gone. For years the table with daily and weekly changes lived on DailyFX, together with the notes of the kind “traders are net-long for the first time since…”. DailyFX closed on 4 September 2024 and its address now opens IG’s own site.

The DailyFX IG client sentiment pages, the PDF guide, the weekly summary — none of that is maintained any more. What is left is the live number per market, and for coders an IG client sentiment API forex pairs included, inside IG’s trading API. If you want forex client sentiment from more than one broker, you have to collect it from each broker’s platform yourself. Nobody publishes the total.

This is the first thing to keep in mind about forex IG client sentiment data in 2026 — it is one broker’s clients, counted by heads, at one moment. Not the market. Not the banks. Not even all retail traders, only the ones who chose IG. If you search for IG EURUSD sentiment, you land on one market page with two numbers on it, and that is the whole dataset.

The crowd and the eighty percent

IG prints on every page that 80% of its retail accounts lose money. ESMA, the EU markets regulator, found in March 2018 that 74% to 89% of retail CFD accounts lose money across EU brokers. Why would a group that loses this reliably be worth watching? Because a losing account is not a random account.

What does a losing trader do when EUR/USD falls 60 pips against his long? He holds. What does he do when it rises 20 pips in his favour? He closes. What does he do when it falls another 60? He adds, because the euro is cheaper now.

On EUR/USD a 100-pip move is $1,000 on one standard lot and $10 on a 0.01 lot, so the trader on a cent account feels the same percentage loss, just in smaller money. The reflex is the same at every size. If you are not sure how a pip turns into dollars, the arithmetic is in my what is a pip article.

This is how a crowd ends up 64% long in a market that has been falling — not because 64 people out of 100 predicted a rise, but because the losers are still in and the winners have already left. The IG sentiment indicator counts open positions. Open positions are mostly the ones under water. If you read client sentiment, you need to read it as a count of trapped positions, not as a forecast.

IG client sentiment vs COT

There is another crowd on the same pair, and it is counted in contracts rather than heads. Every Friday the CFTC publishes what large speculators — funds, CTAs, big trading firms — hold in euro futures as of the previous Tuesday. As of 27 October 2026 they held 230,646 long contracts against 374,540 short, according to the Commitments of Traders report. Net short 143,894.

In the same percentage terms IG uses, that is 38% long and 62% short — 230,646 divided by 605,186 open speculative contracts. On the same pair IG accounts are 64% long. Two groups, one market, opposite sides.

Which group is right? Seven weeks earlier, on 15 September, the funds were short only 26,993 contracts. They added shorts in five of the next six weeks, so the euro has not been giving them a reason to stop. The IG crowd, by the only measure I have — today’s 64% — sits on the other side of that trade.

Most articles treat the funds as the smart side and the retail number as the dumb side. For me it is not that simple. The funds are crowded too — their short side is 43% of all open interest in euro futures — and crowded funds get squeezed like anyone else.

I do not know who is right by the end of November. What I do know is that when both groups lean the same way there is no sentiment trade at all, and that happens more often than the articles admit.

Trend first, crowd second

IG’s own old guide put the line at two to one, not at 60%. Two long for every short is 66.6% long. And it never said “fade 66.6%”. It said: a strong trend, plus the crowd leaning against that trend, gives a signal in the direction of the trend. The crowd is the second half of the sentence, the trend is the first. Most people quote the second half and forget the first.

Put the four rows from my table through this grid. USD/JPY trades near 158.7, higher than in September, and 63% of IG accounts are short it — the crowd is against the trend, and that is the one row that fits without doubt. EUR/USD at 64% long fits only if your daily chart shows a downtrend when you read this. I am writing on 4 November and the chart will have moved, so check it rather than trust me.

GBP/USD at 57% and AUD/USD at 55% are nothing. Fifty-seven is not a crowd, it is a slight lean. With the Bank of England decision at 12:00 GMT on Thursday 5 November it will be a different number by Friday anyway. How the pound reacts to that kind of decision is a separate topic, and I wrote about it in the hawkish vs dovish article.

Two more things belong in the same sentence as the trend. The stop — a sentiment trade is a trade against the people who refuse to use one, so you need to be the one who does. And the size.

The crowd can stay 64% long for three more weeks and the pair can rally 150 pips in between, which is $1,500 on a standard lot and $15 on 0.01. If that number hurts, the position is too big for a signal this soft.

Four ways to lose with it

The first — trading a 55% reading. Half the readings on IG’s pages are between 45% and 55% on any given day, because most markets have no crowd. A number near fifty says that the people long and the people short roughly cancel out. There is nothing to fade.

The second — fading a crowd in a range. In a sideways market the crowd is often right, or at least not wrong for long. Buyers at the bottom of the range are 70% long because the bottom of a range is where you buy. Sentiment without a trend behind it is noise with a percentage sign on it.

The third — forgetting the sample. The number is IG’s clients. Not yours, not your broker’s, not the interbank market, and not the futures market from the chart above. A different broker with a different client base can show 55% long while IG shows 64%. I have no way to check which one is closer to the whole retail world, and neither does anyone who writes about IG market sentiment as if it were the market.

The fourth — chasing the change. “Net-long positions rose 12% since yesterday” reads like momentum. Usually it is the losers adding to a losing long, the reflex from the eighty-percent section above. A rising long count in a falling market is the trapped crowd getting bigger. This is not a reason to join them, and it is not yet a reason to sell either — the trend decides that, not the delta.

Sentiment inside a robot

Forex V Portfolio trades 14 pairs on H1 and has no sentiment input. Not because I decided against it — because there is nothing to test. IG publishes no history of the number. You cannot download a year of EUR/USD readings and run them through the MetaTrader strategy tester next to price, the way I test every filter that goes into Forex V. A filter I cannot backtest does not go into the code.

If you want a sentiment IG trade inside a robot, you need to log the readings yourself — every day, at the same server time, for months — and only then test them. Until then the number is a thing to read with your eyes, once, when the chart already shows a trend. I keep my real account on the YouTube stream and let the robots trade their own rules. Client sentiment stays on a second screen, where it belongs.

Heads, not money. One broker, not the market. A count of trapped positions, not a forecast. Trend first, crowd second. The more of these you keep in view, the steadier the few trades you actually take against the crowd.

On 27 October 2026 the large speculators were short 143,894 euro contracts. On 4 November 2026, 64 out of every 100 IG accounts in EUR/USD were long. The next CFTC report is due on Friday 6 November at 20:30 GMT.

Figures checked on 4 November 2026: CFTC Commitments of Traders, legacy futures-only report, euro FX, positions as of 27 October 2026; ESMA press release of 27 March 2018 on CFD product intervention; Finance Magnates on the closure of DailyFX in September 2024; IG public market pages for EUR/USD, GBP/USD, USD/JPY and AUD/USD.