Do signal channels lean long or short? The split is nearly even

Plenty of followers assume a signal channel is a permabull that only ever says buy. In our archive the direction split is close to even, tilted slightly short. Which means 'they only ever call long' is not the thing costing you money — that sits one step later, in which calls you can actually fill.

Do the channels only ever call long?

Scroll a signal channel through a green week and it can read like a single unbroken cheer. Every post is a long, every chart points up and to the right, every screenshot is a closed winner. It is an easy suspicion to form: the channel is a permabull, it will call buy in any weather, and the only reason a call ever works is that the whole market went up that day. Followers say a version of this constantly — that a channel "just says long and hopes."

It is a fair thing to wonder, and unlike most complaints about signal channels it is checkable. We replay every signal in our index against real exchange candles, which means we also record the direction of every call: long or short. So instead of arguing from a week of scrollback, we can count the whole archive and see which way the calls actually point.

The short version is that the permabull picture does not hold. In our archive, as of August 2026, the direction is close to a coin flip.

What does the archive actually show?

Across the 3,927 signals we have replayed to a resolved outcome, as of August 2026, the split is 2,163 short to 1,764 long. In percentages that is about 55% short and 45% long — tilted toward short, but nowhere near the one-sided "they only ever say buy" that the suspicion describes. If anything the tilt runs the opposite way from what a permabull story would predict.

Two things are worth saying carefully about that number, because it is easy to over-read.

First, it is an aggregate across many channels, not a description of any single one. Some publishers in our index do lean hard one way; a channel built around a single directional thesis will look nothing like the pooled figure. The 55/45 is the shape of the whole corpus, and a corpus is not the channel you are about to pay. We have made this point before about unfillable calls, where a near-even headline number hides the fact that one publisher drives most of the problem while the rest sit far lower. Direction can hide the same way. Always read the split for the specific feed, not for the market of feeds.

Second, "short" and "long" here mean the side the channel named, not a bet on the market's future. A near-even split is exactly what you would expect from feeds that post around the clock across hundreds of symbols. It tells you the calls are not mechanically one-directional. It does not tell you the calls are good.

What does a near-even split actually tell a follower?

This is the part that matters, and it is easy to get backwards.

The useful thing about the 55/45 figure is mostly what it rules out. "They only ever call long" is a specific, testable accusation, and in the aggregate it is false. That closes off one theory of how a channel could be misleading you — the theory that it is a broken clock pointing permanently at buy, right only when the tape happens to agree. The clock is not stuck. Calls go both ways in roughly even measure.

But ruling that out does not put anything in the win column. A follower's instinct is to treat "the direction split looks reasonable" as a point in the channel's favour, and it is not one. Direction is simply not where the edge or the loss lives. A channel can call long and short in perfectly balanced proportion and still be worthless to follow, because knowing which way a call points tells you almost nothing about whether following it made or lost money.

Two channels can post the identical short-to-long ratio and be completely different trades to sit in. The ratio does not capture entry quality, it does not capture whether the stop was survivable, it does not capture whether the call was fillable at all. It is a description of the feed's temperament, not its usefulness. The same trap shows up with headline accuracy: a channel can wave a big win-rate number and still bleed a follower dry, which is why win rate on its own means nothing without knowing the size of the wins against the size of the losses. Direction is one more field that sounds like a verdict and is not.

So the honest reading of the near-even split is narrow: it kills a common suspicion, and it does not hand the channel a compliment. That is genuinely all it does.

Does direction change which calls you can fill?

Here is where the number becomes something a follower can act on, rather than just an interesting stat.

The real cost in following signals, in our data, is not direction — it is execution, and specifically which calls a follower can actually fill. A large share of published signals are overtaken by their own move: price runs to the target before the entry the channel posted ever becomes available, so the trade the channel described was never a trade you could have taken. In our archive that cancelled group is enormous, and we covered its full shape in a separate piece on why half the trades never happen.

The question worth asking is whether that unfillability is a direction artifact — whether, say, the shorts are the ones that never fill, so a follower could dodge the problem by only taking longs. The archive says no. Among the cancelled outcomes, as of August 2026, the split is 998 short to 960 long — about 51% to 49%, even flatter than the resolved population. Unfillable calls are not concentrated on one side of the book. A follower cannot sidestep the execution problem by picking a direction, because the execution problem does not care which way the call points.

That is the point the direction number is actually good for. It closes two doors at once. The channels are not permabulls, and the calls you cannot fill are not hiding on one side. Both of the simple, direction-shaped explanations for why following hurts turn out to be dead ends — which forces the attention onto the thing that is left.

So where does the loss actually live?

Once direction is off the table, the questions that decide a follower's result are all about mechanics.

Can you fill the call at the price posted? This is the single biggest one, given how much of the archive is overtaken before entry. A channel that posts entries you can never actually reach is not giving you its results, it is giving you its highlight reel.

How large is the risk on each call, relative to the reward? Direction tells you nothing here. A feed of balanced longs and shorts with no survivable stop will still take an account apart. This is the same gap that makes a bare accuracy figure meaningless.

Is the feed even one feed? Some of what looks like a diversified spread of calls is a single feed republished under several brands, so a split you think you are averaging across channels is really one channel counted more than once. It is worth knowing what a channel actually posts about at all — in our index the coins channels call are overwhelmingly thin, obscure symbols rather than the majors, and the leverage they attach is usually a fixed channel setting rather than a per-trade decision. Those two fields, like direction, describe the feed's habits. None of them is a verdict on the feed.

The through-line across all of this: the fields that are easy to eyeball — direction, leverage, headline win rate — are the fields that tell a follower the least. They feel like evidence because they are legible. The things that actually move a follower's balance are less visible and take a replay to see: whether the entry was reachable, whether the stop was real, whether the wins were bigger than the losses.

What to do with the direction number

Treat it as a suspicion-killer and nothing more. If you catch yourself thinking a channel is rigged to always say buy, the aggregate says that story is unlikely — calls run close to even, if anything slightly short, as of August 2026. Check the specific channel's own split if you can, because the pooled figure hides publishers that genuinely lean one way. But do not let a reasonable-looking ratio talk you into trusting the feed.

The money you can control is not in the direction of the call. It is in whether you could fill it, whether the risk on it was survivable, and whether the channel's own history holds up when the calls are replayed under one honest rule set instead of read off a screenshot. Direction is the field that everyone stares at and the field that decides the least. Knowing that is worth more than any single long-versus-short count, because it points your attention at the part of following where the loss is actually yours to avoid.