The worst position to be in is not a screen that is red or a screen that is green. It is two screens that both look confident and point in opposite directions. The liquidity composite is grinding higher, central bank balance sheets and reserves are loosening, credit spreads behave. And right next to it the recession model is deteriorating, the yield curve is doing something ugly, employment leads are rolling over. Both dashboards are well built. Both are telling you the truth about the thing they measure. They just cannot both be steering.
I used to treat this as a signal to work harder, to find the one number that broke the tie. That was a mistake. The tie is the information. When a good liquidity read and a good risk read disagree, the honest answer is that the market is in a transition where the two forces have different clocks, and no amount of staring resolves which clock you are on until it is mostly over.
Why the two composites disagree in the first place
They measure different layers of the same system, and those layers move at different speeds. Liquidity is about the supply of money and the willingness to lend it. It is fast, it is a flow, and it tends to lead risk assets because more money looking for a home bids prices up before anything real changes in the economy. Recession models are about the real economy, output, jobs, demand. That is slow, it is a stock of conditions that builds up, and it lags.
So a period where liquidity is rising while recession odds climb is not a contradiction. It is the normal shape of a late cycle. Policy loosens or global reserves swell precisely because the real economy is weakening. The liquidity is the response to the risk. Reading them as opponents when one is partly caused by the other is where people get hurt.
The trap is that both composites will keep printing their view with the same confidence right up until the resolution. A recession model does not get quieter as it gets more wrong in the short run. A liquidity gauge does not warn you that its tailwind is about to be overwhelmed. Neither dashboard knows the other exists.
Which one tends to win, and over what horizon
My rough working rule, and I want to be clear it is a tendency and not a law, is that liquidity wins the near term and risk wins the far term. Over a horizon of weeks to a couple of months, a strong rising liquidity read usually drags asset prices with it even while the recession model deteriorates. Money in the system has to go somewhere, and it front runs the bad news. Over a horizon of quarters, the real economy asserts itself. If earnings and employment actually break, no amount of liquidity holds valuations that were built on the assumption they would not.
Which is why the single most useful question is not which composite is right. It is what is my holding period. If you are trading a two week move, weight the liquidity read heavily and treat the recession model as background. If you are positioning a book you intend to carry for two or three quarters, invert that. Most of the pain I have watched people take came from using a fast signal to justify a slow position, or a slow signal to sit out a fast move that paid.
There is a second wrinkle worth naming. Liquidity tends to win smoothly and lose suddenly. The upside from a liquidity tailwind arrives as a grind, and the downside when risk finally wins arrives as a gap. That asymmetry is the whole reason I do not like being maximally long into a disagreement, even when the near term odds favor it.
Size down instead of picking a side
The practical move when the two composites fight is to reduce exposure, not to resolve the argument. You are being told, correctly, that the distribution of outcomes is wider than usual. The right response to a wider distribution is a smaller position, full stop. Conviction is a function of agreement between independent reads. When they disagree, your conviction should mechanically fall, and so should your size.
Here is the checklist I actually run when the liquidity and risk scorecards diverge.
- Write down my real holding period first, before I look at either composite again. This decides which signal gets the vote.
- Cut position size relative to what I would carry if both agreed. Roughly halving is a sane default when two good composites conflict.
- Keep direction from the signal that matches my horizon, but take the size down from the disagreement.
- Set the invalidation on the signal I am NOT trading. If I am long on liquidity over weeks, my stop is the recession model resolving early, not a price level.
- Watch credit, not equities, for the turn. Spreads usually crack before the index does, and credit is where the liquidity story dies first.
That fourth point is the one people skip. If you are riding the liquidity tailwind, the thing that ends the trade is not the price falling, it is the slow signal arriving faster than expected. So the risk composite you chose not to trade becomes your alarm. You are not ignoring it. You are using it as the stop.
The failure mode to avoid
The classic blowup is to average up into the disagreement because the fast signal keeps being right week after week. Liquidity carries you, you feel smart, you add. The recession model has been wrong for a month so you stop respecting it. Then it resolves in a session and you are large at exactly the wrong moment, because the smooth win talked you into the size that the sudden loss punishes.
The defense is boring. Let the disagreement cap your size, keep the alarm live on the signal you are not trading, and accept that you will leave money on the table in the weeks the fast signal keeps paying. Leaving that money on the table is the premium you pay for not being the person who is maximally long into the gap.
I do not have a clever way to tell you which composite wins in any given episode, and I distrust anyone who says they do. What I have is a way to be smaller and more awake while it plays out, which is usually enough. The tape you cannot read is the tape where you carry less and watch the alarm you set on the signal you chose to ignore.