Larsson Line, MoneyLine & Co.: Thousands of Dollars for a Principle You Can Get for Free

The Larsson Line Pro from CTO Larsson costs $3,500 a year. At Bullmania, the members' club run by Ivan Liljeqvist, better known as Ivan on Tech, membership including the MoneyLine costs $5,000 to $15,000 a year, and you are admitted after an application call.

Neither vendor refunds anything the law does not require. And with both, you see the indicator only after the money is gone.

So you are buying a line you cannot check beforehand. It can also work the other way round: test it yourself first, then decide. In the Backtesting Arena, the platform behind this blog, the rules are open, you turn the settings and you see every trade, across 27 strategies and hundreds of coins. It costs little, and anyone who registers by December 31, 2026 gets two months of Elite as a gift.

That is the test we run here, on eight years of bitcoin prices. That leaves the real question: what are people paying for?

Trend lines on bitcoin are not hocus-pocus

The concession comes first, because it is earned. Moving averages have evidence behind them on bitcoin. Andrew Detzel and co-authors showed in Financial Management in 2021 that the distance between price and its moving averages predicts daily bitcoin returns, including outside the data the model was fitted on.

The test further down also proves the vendors right on one point: every trend line tested made bitcoin's worst crash noticeably smaller.

A vendor who promises that a line gets you out of a bear market is not selling a fantasy. The only open point is whether the paid line does it better than one whose rules are in the open.

What sits behind the two lines

The Larsson Line comes from a man who calls himself CTO Larsson and, by his own biography, spent two decades at Ericsson. The indicator has been on TradingView since June 1, 2020, invite-only. His description: the line is "always in exactly 1 of 3 states – up, down or inconclusive", meant for candles from four hours to one month. Larsson keeps the code hidden. People often search for it as "LarsonLine" or "CTO Line".

Community rebuilds have circulated on GitHub since 2021: four smoothed moving averages (SMMA) on the midpoint of the daily high and low, with lengths 15, 19, 25 and 29. Larsson has never confirmed this formula.

It sits remarkably close to a classic. Bill Williams' Alligator from the 1990s consists of three such smoothed averages on exactly this midpoint, and has shipped free with nearly every charting package for decades.

The MoneyLine is described by Bullmania like this: "It shows you if an asset is in a bullish trend or a bearish trend." It is handed out to members only. On its own blog on October 22, 2026, Bullmania places it as "built on the same ATR-based trend-following concept as Supertrend", plus per-asset tuning and confirmation candles. Bullmania does not give the length or the multiplier.

In the Backtesting Arena, both families are available as open strategies. Every rule is on the strategy page, and you can change every setting yourself:

• EMA Trend Bias for the Larsson Line: two exponential averages and a neutral zone built from volatility (ATR), so again three states: up, down, undecided.
• WMA Trend Signal for the MoneyLine: two weighted averages that cross.
• Plus the Supertrend, which Bullmania itself names as a relative, and the Alligator.

These rebuilds are not 1:1, because the originals are secret. But they belong to the tool families the vendors name themselves or the community has reconstructed. Both paid lines are variants of tools lying around for free. That turns it into a measurable question: is the variant worth the mark-up?

Eight years of bitcoin, four open lines

Bitcoin against Tether, daily candles, January 1, 2018 to November 3, 2026, default settings, same period for all:

All four beat holding on return, and all four get you out of part of the crash. In exchange they sit outside the market 47 to 59% of the time.

At first glance this looks like a ranking, with WMA Trend Signal on top and the Larsson family at the bottom. Does that ranking hold?

On return, no line beats another

The Arena measures how far a strategy's return moves when you turn just one setting by one or two steps, say the length of an average. On daily candles that is up to 17.9 percentage points of return per year.

The largest gap in the table, WMA Trend Signal against holding, is 14.8 points. Between the MoneyLine family and the Larsson family there are 10.1 points. Both sit below what a single knob moves on its own.

Paying for fine-tuning means paying for a difference a turned setting also produces.

The crash tells a different story. Here the threshold is 16.9 points, and three of the four lines clear it comfortably: the Alligator shrinks the worst crash by 37.2 points, the Supertrend by 28.1, WMA Trend Signal by 24.2. The EMA variant of the Larsson family falls just short at 15.2 points.

Trend lines reliably deliver a smaller crash, not more return. So far, though, the whole stretch counts, and it contains two big bull markets. What is left if you look only at the last few years?

Since March 2024, little of the return is left

Neither vendor publishes this test, so we run it here. The Arena splits the stretch: on the first 70% a line may look good, and the verdict falls on the last 30%. Here that is March 9, 2024 to November 3, 2026. Entries fill at the open of the next candle, and all returns are after fees.

The Alligator goes from 53.5% a year to minus 4.8%. Only WMA Trend Signal passes, and even its lead over holding, 4 points before fees, sits far below the 17.9 threshold. The EMA line gets no verdict at all, because its 11 trades are not enough for one.

The two lines standing for the MoneyLine end up furthest apart: WMA Trend Signal and Supertrend are 18.2 points apart before fees, just above the threshold. So the family alone cannot tell you how the original would do.

The crash, by contrast, stays smaller here too: all four sit above holding's −53.0%, with WMA Trend Signal and EMA Trend Bias about 19 points better, past the 16.9 threshold.

The whole stretch shows how well a line fits the past. The last part shows what is left of that.

All four lines are in our own catalog, two fail, and the result is printed here anyway. So the vendors would have to show that their version does more. Do they?

No track record to recompute

Neither vendor has published a list of real trades with date, asset and fees. Larsson's terms say: "Past performance does not guarantee future performance."

The only numbers on the MoneyLine sit in a chart on Bullmania's blog: in 60 of 70 cases price moved one ATR in the right direction after the signal, in 52 cases two, in 50 cases three. Which asset, which period, which costs: not stated. Bullmania writes in the same piece: "Two examples cannot establish consistent outperformance."

Research explains why a curve on show reveals so little. Ryan Sullivan, Allan Timmermann and Halbert White tested 7,846 trading rules on a hundred years of Dow Jones data in 1999. The best rule up to 1986 looked strong; over the following ten years the results were "completely reversed". David McLean and Jeffrey Pontiff found in 2016, across 97 published return signals: 26% less return outside the test data, 58% less after publication.

A curve you cannot recompute shows the vendor's best case, not yours.

Which leaves the question of why so many pay anyway.

What the price convinces people of

In 2008 Francesca Gino gave participants advice from the same adviser, of the same quality, with only the price varying. Paid advice was followed markedly more than free advice, and the more expensive, the more so.

The counterpart came in 2012 from a field study at a large German online bank: around 8,000 active customers were offered independent investment advice for free. Only about 5% took it up, and those who did barely acted on it (Bhattacharya and co-authors, Review of Financial Studies).

Then there is relief. Björn Bartling and Urs Fischbacher showed in 2012 that delegating shifts responsibility effectively, and that this is a strong motive for handing decisions off. Nicola Gennaioli, Andrei Shleifer and Robert Vishny describe in "Money Doctors" (2015) why investors pay fees even though managers trail the market net: trust lowers the risk they feel.

So the price buys commitment and someone to whom the decision can be attributed. That is a real need, but it is no edge on the chart.

What is actually in the packages

In both cases the line is only one part of the package.

Larsson Line Pro includes, per its checkout page, the indicator, automated analytics across 10,000 coins and 100,000 stocks and commodities, unlimited alerts, a 21-episode course and a weekly chart report. No community is listed. The home page says explicitly: "This is NOT a 'copy trading' service or 'trading signal' subscription."

Bullmania lists the MoneyLine, a scanner, an AI tool, a course, private market analysis, livestreams, group coaching and a private community. In a reply on Trustpilot on October 19, 2026, Bullmania writes that new members now need a portfolio of at least $100,000.

So with Larsson you pay for scanner and course, with Bullmania for the inner circle. A community can be worth a lot; it keeps you going and gives you people to talk to. It also passes on mistakes. Rawley Heimer found in 2016 that investors, once given access to a social trading network, became almost twice as prone to selling winners too early and holding losers too long.

An inner circle is a club. No backtest can tell you whether it is worth $15,000 a year. The line alone does not justify the price.

For other well-known paid indicators, the free counterpart is just as easy to find:

"A rebuild is not the original"

The strongest objection: the originals are tuned more finely, so a rebuild that is merely close says nothing about the real Larsson Line.

True, the rebuilds are not identical and the original settings are secret. But that fine-tuning is exactly the knob that moves return by up to 17.9 points. And tuning optimised on past prices loses most of its lead outside those prices, as McLean and Pontiff found.

An original that is better through tuning would have to show it in a track record. Neither vendor publishes one. Untestable does not mean better.

And the original can be checked directly once its signals are available with dates. That takes no formula, only the list. The next section shows how.

See first, then pay

The Arena does not have the better line. The table above shows that: two of our four lines fail.

The difference lies elsewhere. The vendors' line is a black box: you pay, you get the signal, and you have to believe it can do more. The course comes with it, the formula does not.

With us you build the knowledge yourself, for little money and right now at no cost. There are 27 strategies to work with, and the evaluation holds around 143,000 runs on 657 assets.

There you buy an answer you cannot check. Here you learn to find it yourself.

In fairness: the Arena is our product, and we sell a subscription too. So the rule that applies to everyone applies to us: don't take our word for it, recompute it.

What you will not get from us: coaching, livestreams, a club, or buy and sell recommendations. If that is what you want, look elsewhere.

A backtest does not replace a club. It is the check before you spend $3,500 or $15,000.

Before you buy: how to check a paid line

The four lines above are Pro strategies; a free account runs a smaller selection, from 2022. Even so, testing costs you nothing right now: every account registered by December 31, 2026 gets two months of Elite as a gift, with no subscription and no payment details. After that it drops back to Free by itself.

That lets you play everything through in the backtester yourself: every line, every coin, every period, every single trade. What it does not include is API-Pro, meaning the finished verdict from the table above and feeding in someone else's signals. That costs €9.99 a month and comes with paid Elite. Next to $3,500 for the line, that is little.

If you want to carry on, Pro is €9.99 a month and Elite €19.99. The launch offer gives you a full year of Elite for €19.99, and the price then stays locked.

1. Rebuild the line on your coin and your timeframe: EMA Trend Bias, WMA Trend Signal, Supertrend, Alligator. A run in the backtester takes a few seconds. Turn the settings while you are at it and watch how far the result moves. And do not stop at four lines: Strategy Insights shows you which strategy beats holding on which coin.
2. Get the verdict. By hand this costs nothing: run the line in the backtester once up to March 2024 and once from then on, and compare both stretches with holding, which sits right next to the result. If the second stretch collapses, you know enough. The finished verdict from the table above comes from our connector in Claude or ChatGPT, or from the API, and for that you need API-Pro. One question is enough: "Validate EMA Trend Bias on BTCUSDT, daily candles, from 2018."
3. Feed in the real signals. This is the direct route to the original. Ask the vendor for a list with date, asset and fees, from the period after the indicator was released, or ask someone who already has the line for the signals and their dates. The check needs only the date and buy or sell: it fetches the prices itself, fills at the next open and deducts the fees. This also runs through API-Pro. Without a list, all you have is a curve.
4. Watch the trade count. Under 30 closed trades the verdict is always "anecdote". Slow lines feel this most: EMA Trend Bias makes 11 trades on bitcoin in 8.8 years, and 30 would take about 24 years. So run a line like that across several coins.
5. Look at other coins. Bitcoin is a good case for these lines. At 36.8% a year, WMA Trend Signal on bitcoin beats 87.6% of 481 tested pairs, and the median of all pairs is −6.4%. For EMA Trend Bias it is 90.7% of 424 pairs, median −9.8%. Many dead altcoins are in there, which pulls the median down. A line shown to you on bitcoin alone is still the pretty part.
6. Price the extras separately. What would the course, scanner or community be worth to you if the line cost nothing? An open counterpart built on the same principle is already there.
7. Read the refund rules. Larsson refunds only what the law requires (14 days in the EU). Bullmania calls all payments "final and non-refundable", and under its own rules the EU right of withdrawal lapses with the first login.

No study so far compares buyers of such packages directly with non-buyers. What can be measured is the principle behind the line and, with a signal list, the line itself.

Anyone who still buys afterwards at least buys with open eyes. What you learned from the test stays yours, whichever way you decide.

How it is computed

Backtesting Arena, strategy comparison on BTC/USDT, daily candles, 2018-01-01 to 2026-11-03, default settings, one variant per strategy. The table from March 2024 comes from the strategy validation: split after 70% of the stretch on 2024-03-09, fees 0.10% per side, one market, holding as a single purchase at the start of the segment. The thresholds (17.9 points of return, 16.9 points of crash) are measured: that is how far the result moves when a single default setting is shifted by one or two steps. Four lines side by side are a multiple test; the top of the table is therefore more likely overstated.

FAQ

Are the returns after fees? The table shows return before fees. After trading fees each line comes out 0.4 to 1.5 points lower; the gaps do not change.

Is the Larsson Line really four SMMAs of 15, 19, 25 and 29? Only community rebuilds say so; CTO Larsson has not published the formula. The test above does not depend on it: it checks the tool family, not the secret formula.

Are these vendors dodgy? No regulatory proceedings against CTO Larsson or Bullmania are known (search on November 4, 2026). This piece checks what the line can do, not the vendors.

Not investment advice, not a recommendation, not a forecast. Historical patterns are no promise.

Study the Past — Improve your Future 🥋