The BTC regime gate

You can do everything right on a single asset and still lose. The chart looked strong, the entry rule fired cleanly, the stop was placed correctly, and then Bitcoin dropped 12% overnight and every altcoin on the screen went with it. Your analysis was not wrong about that asset. It was wrong about the assumption underneath it, which was that the asset would be allowed to trade on its own merits.

In crypto, most of the time, it is not. Altcoins carry a large amount of shared movement with Bitcoin, and that shared component tends to increase exactly when things get violent. On calm days, individual coins do their own thing. On the days that decide your year, they do not.

Why one instrument gets to veto everything

The trend rules in this track measure an asset against itself. That is the correct question and an incomplete one, because a portfolio of individually valid entries in a falling market is still a portfolio in a falling market.

A regime filter separates "is this asset trending" from "is this market a reasonable place to be adding exposure". Both questions can be answered mechanically, and they answer differently often enough to be worth asking separately.

Indikora uses Bitcoin as the proxy for the second question, tested against a simple moving average on the daily timeframe. Above the average means new crypto entries are permitted. Below it means new crypto entries are blocked.

Using a single asset as the market proxy is a simplification, and it is worth saying so plainly. It works to the extent that crypto correlation holds. In a period where a sector genuinely decouples from Bitcoin, the gate is measuring the wrong thing and will block entries in a market that is fine.

The four settings, and what each one actually chooses

The gate is not one rule. Which moving average it uses depends on the style selected.

Conservative uses the 200-day. Slow, hard to flip, and it will keep you out for months after a major decline, including through the first stretch of a recovery.

Balanced uses the 100-day. Turns back on earlier, and correspondingly turns back on during rallies that fail.

Aggressive uses the 50-day. Reacts quickly, flips state often in a choppy market, and permits entries in conditions the slower settings would still be blocking.

Momentum applies no BTC veto at all. Assets are judged purely on their own trend rules.

There is no ranking here. A longer average is not safer, it is later. The 200-day setting avoids more bad entries and also misses more of the early part of every recovery. The 50-day setting participates earlier and takes more entries that immediately reverse. That is the same tradeoff appearing at four different points on a dial, and which one suits a person depends on things software cannot see.

What the gate does not do

This is the part most people get wrong when they first meet a regime filter.

The gate blocks new entries. It does not close open positions. A position already open when the gate closes is still managed by its own chandelier stop, and it exits when that stop is hit and not before. The regime rule and the exit rule are separate systems, deliberately, because a market-wide condition is not evidence about one particular position.

The gate does not apply outside crypto. Forex, gold and index instruments do not take a Bitcoin veto, because the correlation argument that justifies it does not exist there.

The gate has no view on price. It is not a forecast. It is a statement about whether Bitcoin is currently above or below a line drawn from its own history, and nothing more is claimed.

The cost, stated honestly

Regime filters are easy to admire in a backtest and hard to live with. Any test of this gate measures how many entries it removed and what happened to those entries afterward, over a specific historical window with a specific set of assets. That is useful for understanding behavior and it is not a prediction, because the sample of crypto market cycles is small and every one of them was structurally different from the last.

The lived costs are these. You will sit out the sharpest part of at least some recoveries, because the bottom happens while Bitcoin is still below any of these averages. You will be whipsawed in ranges, where the gate opens and closes repeatedly and each state change costs you a couple of entries. And on the day a decoupled sector runs without Bitcoin, you will watch it from the sidelines with a filter that was measuring the wrong thing.

The reason to accept that is not that the filter is clever. It is that in the specific case that ends accounts, a broad simultaneous decline across correlated assets, a per-asset trend rule offers no protection at all and a market-level rule offers some. You are paying for one scenario with performance in several others, and knowing which one you bought is the whole point of understanding the mechanism.