Five open source tools, one method behind all of them

The job was done by hand for a long time first

The method came out of a portfolio course, and the way it is normally taught is a spreadsheet. You open a chart, add a dozen indicators, look at each one, type up or down into a cell, and read the average at the bottom. Do that for five markets and it is most of an hour, every single day, and every one of those readings is a chance to type the wrong thing or to talk yourself into a vote you wanted rather than the one on screen.

These scripts are that spreadsheet moved onto the chart. Same twelve measurements, same settings, same arithmetic, except the chart does the reading and the typing. The number in the corner of the chart is the number that would have gone into the cell. Nothing is optimised, nothing is predicted, and no part of it can be nudged by how the day feels.

They are published for two reasons. A system nobody can inspect is a system nobody should trust, and a claim about a rule is worth very little next to the rule itself. Everything the ladder does to decide where money sits is now something you can put on your own chart and check against your own data.

Twelve opinions, one number

An indicator is a small calculation that looks at recent prices and says something about them. Some of them follow the trend itself, the way a moving average does, and answer the question is price above its own recent average or below it. Others measure momentum, which is whether a move is still picking up speed or quietly running out of it. Each kind is useful and each kind is regularly wrong, in its own particular way.

So none of them gets to decide on its own. Twelve of them look at the same daily chart. Six follow the trend, six measure momentum. Each one casts a single vote, up or down, with no half votes and no weighting. The twelve votes are averaged and that average is the score.

Because twelve whole votes are averaged, the score can only land on thirteen values, one sixth apart. That makes it easy to read backwards. +0.67 means ten agreed and two did not. +0.17 is the weakest possible yes, seven against five. The size of the number tells you how united the panel is, which is worth knowing, but only the sign changes what the system does.

What a confirmed close is, and what repainting is

A daily candle is not finished until the day is over. Until then its close keeps moving, so anything calculated from that close keeps moving too. An indicator that reads from the unfinished candle will happily show you a buy signal at lunchtime and have erased it by midnight, leaving no trace that it was ever there. That erasing is called repainting, and it is the single most common way a backtest flatters an indicator that never worked.

These scripts do not read from the unfinished candle at all. While today is still running they show yesterday's confirmed reading and nothing else. The score can only change at the moment a daily candle closes, which for crypto and gold is 00:00 UTC. What you see on the chart at any moment is a decision that was already final before you looked at it.

The practical rule falls straight out of that. A flip belongs to the close that produced it, and the earliest honest place to act on it is the open of the next candle. Acting before then means acting on a number that can still change, which is not a system, it is a guess with extra steps. The timing page goes through what that looks like in practice.

Everything on screen, one piece at a time

Adding the script puts four things on your chart. The candles change colour with the score. A triangle with a word beside it marks each flip. A pane underneath draws the score itself as a stepped line between minus one and plus one, filled to the zero line. And a small table sits in the corner with the full breakdown.

There is also a pair of plain moving averages on the chart, the 12 and the 21. They are not part of the vote and they never touch the score. They are there as a sanity check, because the settings were tuned against that cross in the first place, and the table reports how often the two agree.

The shaded vertical bands you will see on the charts below are the days when the panel and that simple cross disagreed. They are worth looking at, because they are the days the extra eleven indicators actually earned their place, for better or worse.

One indicator per rung, five in total

They share a layout and a method and nothing else. Each is tuned for its own market, which means the twelve components differ and the settings differ. Putting the gold one on a crypto chart will produce a number, and that number will be meaningless.

Each has its own colour pair so that five charts stay apart at a glance. Where the question is which of two things is winning, the colours name the two things rather than saying bullish or bearish, because there is no such thing as a bullish ratio.

Rung one

CRYPTOCAP:TOTALES, daily

Confirmed Close MTPI TOTALES

Is the crypto market as a whole trending up or down?

A positive score means: the total value of the crypto market, stablecoins excluded, is in an uptrend.

uptrenddowntrend

Open on TradingView

Rung two

BINANCE:ETHBTC, daily

Confirmed Close MTPI ETH/BTC

When crypto is trending up, which of the two majors is leading?

A positive score means: ether is beating bitcoin. Both can still be falling in dollars, this only measures one against the other.

ETH leadsBTC leads

Open on TradingView

Rung three

BYBIT:PAXGUSDT.P, daily

Confirmed Close MTPI GOLD

When crypto is trending down, is gold trending up?

A positive score means: gold is in an uptrend and is worth holding while crypto is not.

uptrenddowntrend

Open on TradingView

Rung four

SPCFD:SPX, daily

Confirmed Close MTPI SPX

When gold is not trending either, are equities?

A positive score means: the S and P 500 is in an uptrend.

uptrenddowntrend

Open on TradingView

Rung five

TICKMILL:EURUSD, daily

Confirmed Close MTPI EUR/USD

When nothing is trending, which currency should the cash sit in?

A positive score means: the euro is the stronger of the two, so cash sits in euro rather than dollar.

EUR strongerUSD stronger

Open on TradingView

All five sit on the same TradingView profile. They are free and open source. You can read every line, copy it, change it, and check any claim on this page against your own chart.

The twelve components, market by market

Nine of the twelve appear in every market. Two slots swap, one in the trend six and one in the momentum six, because some measurements simply do not behave on some series, and the euro pair swaps a third. The numbers below are read straight out of the published source, not from memory.

One setting looks like a mistake and is not. The Optimized Trend Tracker runs on a weighted moving average for the crypto market and on an exponential one everywhere else, and on the euro pair the Coppock Curve is configured upside down compared with the textbook. Both are what the tuned system actually uses, and both are published as they are rather than tidied up.

Crypto market cap

ETH against BTC

Gold

S and P 500

Euro against dollar

Every script was checked twice before it went out

Seven of the twelve components are rewritten from published open source formulas rather than called in, because a Pine script cannot read another script's output. A rewrite that is subtly wrong is worse than useless, so each one was tested rather than trusted.

First check, against the originals. The script and the twelve real indicators were put on one chart and compared vote by vote across every bar the feed would load. Not the final score, each of the twelve individually.

Second check, against the live record. The score was compared against the daily history the system has actually been run on, which is kept as a row per day and never edited after the fact.

Two of those rows deserve to be read rather than skipped. The S and P 500 column has only twenty days of live record behind it, because that rung was added to the ladder recently, and twenty days is a weak check next to six hundred. The euro pair has one day out of six hundred and twenty six that does not match, and the cause is known: euro dollar candles close at 21:00 UTC rather than midnight, so a reading taken earlier in the evening records the previous session. It costs at most a day of lag, on the last rung, which the ladder only ever reaches when everything above it is already negative.

There is one caveat inside the dashboard itself. The agreement percentage is measured against the raw 12 and 21 cross, while the tuning used a cleaned version of that cross with every regime shorter than eight days merged away as noise. So the figure on screen is a little lower than the one the tuning was scored on. It is reported that way on purpose rather than quietly improved.

From a blank chart to a reading you can act on

1. 01Open a free TradingView account if you do not have one. Nothing here needs a paid plan.
2. 02Open the script you want from the cards above and press Add to favourite indicators. It then appears under Favourites in the Indicators menu on any chart.
3. 03Set the chart to the exact symbol on the card and to the daily timeframe. Both matter. A four hour chart of the same symbol produces a completely different and completely untested number.
4. 04Add the indicator. Leave every setting alone. The defaults are the tuned values, and changing one of them means you are running something nobody has checked.
5. 05Read the corner table. The number and the word beside it are the answer for that rung. Then go to the ladder and see which question you are meant to ask next.

Getting told instead of looking

Each script carries two alerts, one for a turn up and one for a turn down. Right click the chart, choose Add alert, pick the indicator as the condition and then the direction you want. Set the trigger to Once per bar close. Anything else will fire you out of bed for a number that has not finished forming.

An alert is not an instruction to trade immediately. It tells you a close has flipped a rung. The trade belongs on the next open.

Most of the parts are other people's work

Nothing in the twelve was invented here. The components are either textbook indicators or open source community scripts, reimplemented so that one script can read all of them at once. The authors of the community ones are named in every publication and are worth following in their own right.

What is original is the small part: putting twelve of them behind one number, holding that number until the candle closes, showing the breakdown, and wiring five of these into a ladder that decides where the money sits.

The honest limits, stated once and plainly

They are always late. A trend has to break before twelve indicators notice, so every exit is below the high and every entry is above the low. That is not a flaw waiting to be tuned out. It is what you pay for only acting on finished information.

They get chopped up sideways. In a market that goes nowhere for months the panel flips back and forth, selling and buying back a few percent higher. The regime counter in the corner is the warning light for this: a string of short regimes means the tool is not working on that market right now.

A daily close cannot save you from a gap. An index or a currency can open several percent below where it closed, and no daily indicator has anything to say about that. The crypto rungs run around the clock so they suffer less from this, and they make up for it by moving further.

Past behaviour proves nothing. Every check on this page is a check that the code computes what it says it computes. It is not evidence that the method will make money. Those are different claims and only the first one has been demonstrated.