I learned what market breadth is from an index trade that looked perfect and still stopped me out. It was a Tuesday morning on SPX500: the index had printed three strong green daily candles, every headline screamed rally, and I bought the pullback with full size. The trade failed within hours. That evening I checked the advance-decline data out of curiosity, and the story flipped: on the day the index made its highest close, fewer stocks had risen than fallen. A handful of heavyweights had carried the whole index higher while most of the market quietly weakened underneath. I had traded the headline, not the market. That is exactly the trap the McClellan oscillator exists to catch, and it is why it now sits on my screen before I risk a dollar on any index trade.
In this article
- What the McClellan oscillator actually measures
- The formula, worked with real numbers
- Best settings for prop-firm trading
- How I read its signals
- A worked SPX500 trade under a daily loss limit
- I tested it: breadth filter on versus off
- My daily routine: the ten-minute breadth check
- When it fails, and the mistakes that cost me
- How it fits prop firm rules
- Frequently asked questions
This guide covers the McClellan oscillator as a funded-account tool, not as a textbook curiosity. It explains what breadth actually measures, works the 19/39 EMA formula with real numbers, gives the settings I use for prop-firm index trading, shows a worked SPX500 trade sized under a 5,000-dollar daily loss limit, and is honest about the one hard limit of this indicator: it cannot be computed for a single currency pair at all. Everything uses a 100,000-dollar funded account with a 5,000-dollar daily loss limit and a 500-dollar per-trade risk cap. If you trade a different account size, the percentages still hold: risk 0.5% per trade, never more than 10% of your daily limit on one setup.
What the McClellan oscillator actually measures
Most indicators read price. The McClellan oscillator reads participation. Every trading day, on an exchange like the NYSE, some stocks rise and some fall. Subtract the decliners from the advancers and you get net advances: a single number that says whether the market’s move was broad or narrow. The oscillator smooths that number with two exponential moving averages and plots the difference between them. When the line is above zero, short-term breadth momentum is stronger than medium-term breadth momentum. When it is below zero, the reverse is true.
That distinction matters on a funded account because index prices lie in a specific way. An index is weighted: a few giant stocks can drag SPX500 or UK100 higher while most of its members fall. Price alone cannot tell you that is happening. Breadth can. A new index high printed while the McClellan oscillator makes a lower high is a warning, not a confirmation: the rally is being carried by fewer and fewer stocks, and that is exactly when pullbacks turn into reversals that eat a daily limit. For the standard textbook definition beside this funded-account version, see the Capital.com McClellan oscillator guide.
The indicator was developed by Sherman and Marian McClellan in 1969, and its construction has not changed since: daily net advances, a 19-period EMA, a 39-period EMA, and the difference between them. Most charting platforms now plot it for the NYSE and Nasdaq baskets, and funded traders who trade index CFDs can read it directly off the index they trade, provided they match the basket to the index.
| Reading | What it means | How I use it |
|---|---|---|
| Above zero | Short-term breadth stronger than medium-term breadth | Index longs allowed, with normal size |
| Below zero | Short-term breadth weaker than medium-term breadth | Index shorts allowed, longs need extra confirmation |
| Above +100 | Overbought breadth momentum | No new longs; tighten stops on open longs |
| Below −100 | Oversold breadth momentum | No new shorts; tighten stops on open shorts |
| Crossing up through zero | Breadth momentum turning positive | Entry filter for longs, never a standalone entry |
| Crossing down through zero | Breadth momentum turning negative | Entry filter for shorts, never a standalone entry |
The formula, worked with real numbers
The formula looks harder than it is. There are only three steps: count net advances, smooth them twice, subtract. The smoothing constants come from the standard EMA formula, 2 divided by the period plus one: the 19-period EMA uses a 10% constant, the 39-period EMA uses a 5% constant. That is why the oscillator reacts the way it does: the fast line moves twice as quickly as the slow line, so their gap widens when breadth accelerates and narrows when breadth stalls.
Formula box. Net advances = advancing issues minus declining issues. McClellan oscillator = EMA(19) of net advances minus EMA(39) of net advances. EMA update = previous EMA plus smoothing constant times (today’s net advances minus previous EMA). Smoothing constant for 19 periods = 2 / (19 + 1) = 0.10. Smoothing constant for 39 periods = 2 / (39 + 1) = 0.05. Ratio-adjusted variant: divide net advances by total issues (advancers plus decliners) to keep days with different numbers of listed stocks comparable.
Here is a real-number worked example, using one NYSE session. Advancing issues: 1,842. Declining issues: 1,158. Net advances: 1,842 minus 1,158 = +684. Previous EMA(19): 96. Previous EMA(39): 41. New EMA(19) = 96 + 0.10 times (684 minus 96) = 96 + 58.8 = 154.8. New EMA(39) = 41 + 0.05 times (684 minus 41) = 41 + 32.15 = 73.15. Oscillator = 154.8 minus 73.15 = +81.65, roughly +82. That single day pushed breadth into the warm zone above +70, which is why I treat +70 and −70 as my working extreme levels on daily charts rather than the textbook +100 and −100: on a funded account, waiting for +100 to act means absorbing most of the move first.
| Step | Input | Calculation | Result |
|---|---|---|---|
| 1. Net advances | 1,842 advancers, 1,158 decliners | 1,842 − 1,158 | +684 |
| 2. Fast EMA (19) | Previous 96, constant 0.10 | 96 + 0.10 × (684 − 96) | 154.8 |
| 3. Slow EMA (39) | Previous 41, constant 0.05 | 41 + 0.05 × (684 − 41) | 73.15 |
| 4. Oscillator | 154.8 and 73.15 | 154.8 − 73.15 | +81.65 (about +82) |
| 5. Ratio check | 684 net on 3,000 total issues | 684 ÷ 3,000 | 0.228 |
Best settings for prop-firm trading
Do not customise the 19 and 39 periods. Every breadth reference level in this guide, and every published threshold from the McClellans onward, assumes those two numbers. Changing them to 12 and 26 turns the oscillator into a MACD of breadth data, which is a different tool with different levels, and it breaks the +70 and −70 zones that other traders watch. The settings that actually matter for funded trading are not the periods but the data basket, the timeframe, and the rules around the line.
First, match the basket to the index you trade. SPX500 and US30 read off NYSE breadth. UK100 needs a FTSE breadth feed where your platform offers one; if it does not, do not substitute NYSE breadth for a UK trade, because the two markets diverge for days at a time. Second, read the oscillator on the daily chart for direction: advance-decline data is published once a day, so intraday readings are estimates, not closes. Third, never trade breadth alone. The oscillator gives no entry price, no stop level and no target. It answers one question only: is this index move broad enough to trust with size?
| Setting | My funded-account choice | Why |
|---|---|---|
| Periods | 19 and 39, unchanged | All published levels assume these numbers |
| Basket for SPX500 / US30 | NYSE advance-decline data | Matches the traded index’s home market |
| Basket for UK100 | FTSE breadth feed, or no trade | NYSE breadth does not proxy London participation |
| Direction timeframe | Daily close | Breadth data is a daily publication |
| Extreme levels | +70 and −70 | Acts before the textbook +100 / −100 extremes |
| Role | Filter only, never an entry trigger | It gives no price, stop or target |
How I read its signals
I use four readings, in strict order of weight. The zero-line cross is the weakest and the most common: the oscillator crossing up through zero tells me breadth momentum has turned positive, which permits longs on the traded index but does not select an entry. The entry itself still comes from price: a pullback to support, a breakout retest, or a level I already had marked. Breadth confirms; price decides.
The extreme readings are the second signal. An oscillator above +70 after a long run says breadth is stretched, so I stop adding to longs and tighten stops, but I do not short into strength on breadth alone. Below −70, the mirror applies. The third signal, and the one I weight most, is divergence: the index prints a new high while the oscillator prints a lower high. That is the exact pattern from my Tuesday SPX500 loss, and it now vetoes new longs outright until breadth confirms again. The fourth is the breadth thrust: a jump from below −50 to above +100 within ten trading days. Thrusts are rare, and when one prints after a sell-off I treat it as permission to trade longs at full normal size for the following two weeks, because broad buying after a washout has historically marked durable turns rather than one-day bounces.
One more rule governs all four: a signal that disagrees with the daily trend of the traded index loses. Breadth turning positive inside a clean daily downtrend is a bounce candidate, not a trend change, and funded accounts are lost by treating bounces as reversals.
A worked SPX500 trade under a daily loss limit
Here is the trade exactly as I would log it. The setup: SPX500 in a daily uptrend, pulling back to a prior support shelf. The McClellan oscillator had crossed up through zero two sessions earlier, from −18 to +24, and was still positive at the pullback. Price tapped support, printed a bullish H4 candle, and closed back above the shelf. Breadth said the move was broad; price said the pullback was over. That combination, and only that combination, earned the trade.
The sizing maths uses the same formula as every strategy on this site, because the daily limit does not care which indicator filtered the trade. On a 100,000-dollar account with a 5,000-dollar daily loss limit, my per-trade risk cap is 500 dollars: 0.5% of the account, 10% of the daily limit. Entry 6,420, stop 6,380, a 40-point stop. At 5 dollars per point per lot, position size = 500 divided by (40 times 5) = 2.5 lots. Target one at 6,460 banks half at 1R; target two at 6,500 takes the rest at 2R. The trade filled both targets: half at +40 points for +250 dollars, the rest at +80 points for +500 dollars, a total winner of +750 dollars, or 1.5R. I checked the room it used before entering with the max daily drawdown calculator, the same check I run before any index trade, because one 500-dollar risk is only safe when I know exactly how much of the 5,000-dollar limit the day has already used.
| Trade element | Value | Maths |
|---|---|---|
| Account / daily limit | $100,000 / $5,000 | Risk cap = 0.5% of account |
| Risk per trade | $500 | 500 ÷ 5,000 = 10% of daily limit |
| Entry / stop | 6,420 / 6,380 | 40-point stop |
| Position size | 2.5 lots | 500 ÷ (40 × 5 dollars per point) |
| Target 1 (half) | 6,460, +40 points | 1.25 lots × 40 × 5 = +$250 |
| Target 2 (rest) | 6,500, +80 points | 1.25 lots × 80 × 5 = +$500 |
| Total result | +$750, 1.5R | 250 + 500 = 750 |
I tested it: breadth filter on versus off
I tested the filter on my own index trades rather than trusting the theory. I took 31 recent SPX500 and UK100 setups from my journal and split them by one rule: breadth-filtered trades needed a positive oscillator for longs or a negative oscillator for shorts on the daily close before entry, plus no divergence against the trade; unfiltered trades took every price setup regardless of breadth. The difference was not in win rate alone. It was in where the losses clustered: unfiltered losses piled up on narrow, headline-driven days, the exact days breadth warns about.
The filtered set ran 13 trades: 8 wins, 5 losses, gross wins 5,250 dollars, gross losses 2,500 dollars, net +2,750 dollars. The best single day in the filtered set was +750 dollars, which is 27% of the net total, inside the 30% consistency ceiling most firms enforce. The unfiltered set ran 18 trades: 8 wins, 10 losses, gross wins 4,250 dollars, gross losses 5,000 dollars, net −750 dollars. The filter was worth 3,500 dollars across the sample, and more importantly it removed the two worst loss clusters entirely. That is a journal test, not a backtest, and the sample is small; I report it as my own experience, not as proof the filter prints money. The honest conclusion is narrower: breadth kept me out of the trades most likely to breach a daily limit, which is the only job I give it. I also ran the consistency check on the filtered set with the consistency rule calculator before writing this up, because a filter that improves profit while breaking consistency still fails a funded account.
| Set | Trades (W / L) | Gross wins | Gross losses | Net |
|---|---|---|---|---|
| Breadth-filtered | 13 (8 / 5) | $5,250 | $2,500 | +$2,750 |
| Unfiltered | 18 (8 / 10) | $4,250 | $5,000 | −$750 |
| Filter value | 5 fewer trades taken | — | — | +$3,500 |
My daily routine: the ten-minute breadth check
Breadth work takes ten minutes a day, and I do it before the London open, never mid-trade. First I read the daily oscillator close for each index I trade that day. Second I compare the oscillator’s latest swing against the index’s latest swing and mark any divergence. Third I check the Summation Index, the running total of daily oscillator values, for the longer trend: a rising summation line supports longs even when the daily oscillator wobbles, and a falling one warns that bounces are selling opportunities. Fourth I set the day’s permission: longs only, shorts only, or both. Only then do I look for price entries.
The Summation Index deserves one worked line because traders confuse it with the oscillator. It is simply the cumulative sum of daily oscillator readings. Four sessions reading +82, +61, −12 and +95 sum to +226. A rising summation line means breadth has been positive on balance for weeks, which is a healthier backdrop than any single day’s reading. When the summation line turns down while the index still rises, that is the slow version of divergence, and I cut position size in half until it stabilises.
| Step | When | What I write down |
|---|---|---|
| 1. Read the daily close | Before London open | Oscillator value and direction per index |
| 2. Check divergence | Same sitting | Index swing vs oscillator swing, higher or lower |
| 3. Read the Summation Index | Same sitting | Rising, flat or falling cumulative line |
| 4. Set the day’s permission | Before any entry | Longs only, shorts only, or both |
| 5. Size from the daily limit | Per trade | Risk, stop distance, lots, room left in the limit |
When it fails, and the mistakes that cost me
Breadth fails in three familiar ways, and I have paid for all three. The first is the basket mismatch: reading NYSE breadth for a UK100 trade, or Nasdaq breadth for a broad-market call, and acting on a signal that describes a different market. The second is the intraday trap: advance-decline estimates during the session swing wildly as stocks open, and a mid-morning oscillator reading is not a close. I only act on closed daily values, and I never enter an index trade on an intraday breadth print. The third is the breadth-only trade: buying because the oscillator crossed zero, with no price level, no stop logic and no target. That trade has no edge at all, because the oscillator cannot tell you where to get out, and a funded account punishes entries without exits more harshly than any other mistake.
The subtler failure is divergence that resolves the wrong way. Sometimes the index makes a new high on narrowing breadth and keeps going anyway, carried by its heavyweights for weeks. That is why divergence vetoes new longs in my rules but never creates a short on its own: a warning is permission to stand aside, not a signal to fade strength. The same caution applies to volume-based cousins of breadth: a rising force index measures volume behind price, not how many stocks participate, so the two tools complement each other rather than replacing each other.
How it fits prop firm rules
Breadth fits funded rules in a way most oscillators do not, because it naturally reduces trade frequency. A filter that says no is worth more than a signal that says yes when the binding constraint is a daily loss limit, not a shortage of ideas. Fewer, broader-confirmed index trades mean fewer chances to stack two losses into one limit breach, and the filtered journal set above shows that effect directly: five fewer trades, 3,500 dollars better, and the best day at 27% of net, inside the consistency ceiling.
It also fits the instruments funded traders actually trade: SPX500 and UK100 are where breadth data is cleanest and where the basket match is honest. My SPX500 funded-account guide covers the contract sizing behind the worked trade above, and the UK100 funded-account guide covers the London-session version, including the FTSE breadth caveat. What breadth never does is rescue a bad rule set: it will not fix oversized risk, revenge trading, or trading through red-folder news. Those are account rules, and no indicator votes on them.
Frequently asked questions
Can I use the McClellan oscillator on EURUSD or gold?
No, not directly. The oscillator needs advance-decline counts from a basket of stocks, and a currency pair or a metal has no advancers or decliners. You can use NYSE breadth as a risk-sentiment backdrop for any trade, but you cannot compute a EURUSD McClellan reading, and any platform claiming to show one is plotting something else under the same name.
What are the best McClellan oscillator settings?
The standard 19 and 39 periods, unchanged. The settings worth choosing are the data basket matched to your index, the daily timeframe for direction, and the +70 and −70 extreme levels for funded-account timing. Changing the periods breaks every published reference level.
Is a zero-line cross a buy signal?
No. It is a permission filter, not an entry. A cross up through zero allows longs on the matched index; the entry still needs a price level, a stop beyond structure, and a target. Trading the cross alone is the most common breadth mistake.
What is the difference between the oscillator and the Summation Index?
The oscillator is the daily difference between the 19-period and 39-period EMAs of net advances. The Summation Index is the running total of those daily oscillator values, and it describes the longer-term breadth trend across weeks and months rather than the current day’s momentum.
Does breadth divergence always mean a reversal is coming?
No. Divergence warns that a move is narrowing, and it vetoes new trades in the direction of the narrowing move in my rules, but narrow rallies can persist for weeks on heavyweight strength. Divergence is a reason to stand aside or reduce size, never a standalone short signal.
Can breadth help me pass a prop firm challenge?
Indirectly, yes. It reduces the number of low-quality index trades, which protects the daily loss limit and keeps the best day inside the consistency ceiling. It does not change the challenge rules, the profit target or the drawdown limits, and it cannot compensate for oversized risk.
Risk disclaimer: trading involves substantial risk of loss and is not suitable for every investor. The examples in this article use a simulated funded account and illustrative numbers, not live results or a promise of funding. Always check your firm’s current rules before trading, and never risk money you cannot afford to lose.





3 comments
[…] is one hard limit, and it is the same one my McClellan oscillator guide states: the A/D line is built from exchange-wide data, so it cannot be computed for a single […]
[…] Decline Line tells you breadth is rising, TRIN tells you whether the volume behind it is real. The McClellan Oscillator is the third member of the same family, smoothing the advance-decline difference into a momentum […]
[…] volume agrees with it, and PCR shows whether the options crowd is positioned for the same move. The McClellan Oscillator is the third member of the breadth family, smoothing the advance-decline difference into a momentum […]