The first time I built a relative strength grid I made it worse for myself, not better. I had the same list of assets ranked on daily, weekly, and monthly RS, three columns side by side, and I treated each column like it was voting. Whatever sat at the top of the daily column got my attention, because that was the thing moving right now. A few weeks of that and I noticed I was almost always buying the asset that had just finished its move and selling the one about to start.
The grid was fine. My reading of it was the problem. A multi-timeframe RS grid is not three signals to average. It is one object with a fast axis and a slow axis, and those two axes answer different questions. Once I started treating them as different questions the conflicts stopped being noise and started being the most useful thing on the screen.
What each timeframe is actually telling you
Relative strength here just means ranking an asset against a peer set or a benchmark over some lookback, so you get an ordering rather than a raw return. The lookback is the whole game. A daily RS rank built on a short lookback is measuring who is winning the current push. A monthly rank on a long lookback is measuring who has been quietly accumulating strength for a while. These are not the same information at different resolutions. They are genuinely different facts.
The way I hold it in my head:
- The monthly column tells you the regime. Is this asset in the group that has been leading, lagging, or drifting sideways over a meaningful stretch. This changes slowly, which is exactly why it is trustworthy.
- The weekly column tells you the current leg. It is where a real trend change shows up first without being pure noise. If weekly is turning up under a strong monthly, something is starting.
- The daily column tells you timing and nothing else. It is the most seductive and the least durable. A top daily rank can evaporate in two sessions.
So the slow timeframes are for deciding whether you want to be involved at all, and the fast one is for deciding when. If you flip that, if you use the daily to decide what to own and the monthly to time it, you get whipsawed and you deserve to. You are asking the noisy axis to make the durable decision.
Reading the grid as one picture
Here is the workflow I actually use, and it is boring on purpose. I read the columns slow to fast, right to left in my mental layout, and I do not let the daily rank speak until the slower ones have.
- Start monthly. Sort by it. Anything in the bottom of this ranking is off the table for a long trade no matter how good it looks today. I am only shopping in the top portion of the monthly column.
- Inside that shortlist, look at weekly. I want weekly to agree with monthly or to be improving toward it. A name that is strong monthly and rising weekly is a trend that is still intact and getting confirmation.
- Only now look at daily, and only as a trigger. If monthly and weekly line up, a firming daily rank is my entry window. If daily is hot but the two above it are soft, I do nothing.
The mental shift is that the daily rank is never a reason to act on its own. It is permission to act on a decision the slower timeframes already made. When I explain this to people the pushback is always the same, that you will miss the fast movers. You will. Most fast movers on the daily axis are asset that has already run, and the ones that matter almost always show up on the weekly a little later with more room left. Missing the first two days of a real trend has never once hurt me. Catching the last two days of a fake one has.
The disagreement patterns that mean wait
Conflict between timeframes is information, and specific shapes of conflict have specific meanings. These are the ones I have learned to respect.
Hot daily under weak monthly and weekly. This is the classic trap and the one that used to get me. Top of the daily column, bottom of the monthly. This is almost always a countertrend bounce in a lagging asset, a dead cat with good manners. The correct action is to wait, because you are looking at strength that has no structural support underneath it. If it were real, the weekly would start catching up. Watch for that instead of chasing the daily.
Strong monthly, collapsing weekly. A former leader where the slow axis still looks great but the weekly rank has rolled over. The monthly is a lagging photograph of strength that may already be leaving. I do not short this and I do not buy the dip yet. I wait to see whether weekly stabilizes and turns back up, which would be a genuine continuation, or keeps falling and eventually drags the monthly with it, which is a top forming in slow motion.
Everything agrees and everything is already high. Monthly, weekly, and daily all top-ranked at once feels like the best possible setup and is often the worst entry. Full agreement across every timeframe usually means the move is mature and crowded. I will hold something like this if I already own it, but starting a new position when all three axes are maxed out is buying the part of the trend with the least remaining fuel.
Choppy daily, flat weekly, flat monthly. The daily rank is bouncing around the middle and the slower columns are not committing to anything. There is no trend to read here in any direction. This is the grid telling you to go find something else, and the discipline is to actually listen instead of manufacturing a trade out of a daily wiggle.
The through-line is that whenever the fast axis and the slow axis point different directions, the default is wait, not act. The only clean go signals are when the slower timeframes have already made the call and the daily is simply confirming it. Everything else is either a trap dressed as opportunity or a trend that has not decided yet.
A couple of rules that keep me honest
I keep the slower timeframe as a veto, never as a suggestion. If monthly says no, the trade is dead, full stop, regardless of how clean the daily looks. Giving the slow axis veto power over the fast one is the entire point of building the grid.
I also try to name what the daily rank is doing before I look at it, based on the two columns above it. If monthly and weekly are strong, I expect the daily to be a timing tool. If they are weak, I expect the daily to be a trap and I read it defensively. Forming that expectation first stops the daily number from hijacking my attention, which is what it is built to do.
None of this makes the grid predict anything. It just stops me from taking the one signal that is loudest and treating it like the one that is truest. Those are rarely the same column, and the whole skill is in not confusing them.