Weekly Engine #88

W3 March 2026

GM/GN.

This week we released SQL of the Week 017 with a new model - the Distressed Inflow Pressure Index (DIPI).

This model very accurately shows how overloaded exchanges are with loss-making coins. For me, this is one of the key discoveries of the year: DIPI highlights obvious structural insights that often remain hidden in other on-chain models.

The Loss Ratio metric is especially important - it reflects how large the share of loss-making coins flowing into exchanges is at any given moment. It helps see not just the fact of pressure, but its depth and scale. I recommend checking out the issue, bookmarking the model since it is available online on the CryptoQuant platform and adding it to your working set of signals.

Earlier this week, during our morning zoom, we discussed Weekly Engine and the tasks that need to be addressed in the next update. In the process, it became clear: the main priority right now is not expanding functionality. The problem is deeper - the current Weekly Engine architecture and its entire signal set are primarily tailored for long-term investors and analysts. This is a strong foundation, but this is exactly where we started losing focus on traders and more active market participants.

In other words, the question is no longer what else to add to the system, but who it actually works best for right now. And at this stage it became clear: if we want to take the next step, Weekly Engine needs to better cover scenarios not only for long-term analysis, but also for active decision-making. So today I am publishing the first screenshots of the DECISION TERMINAL built on the Weekly Engine framework designed for active investors and traders.

The terminal combines more than ten key Bitcoin on-chain and derivatives metrics - currently 14 signals - into a single composite indicator that shows how suitable the market is for a trade right now. Instead of the usual set of scattered charts and manual monitoring of dozens of indicators, we get a ready-made market interpretation: Market Quality Score (MQS), Execution Window Score (EWS), cycle phase according to Weekly Engine, market direction, and the final verdict - YES / CAUTION / NO - with a clear explanation in plain language.

The terminal’s purpose: to help quickly answer three key questions - is the market environment quality good right now, is execution working, does it make sense to open a position at all. And the key differentiator of the terminal is the built-in LLM model that interprets the current market state and translates a complex set of signals into a clear analytical conclusion without the need to manually parse each indicator.

The terminal updates 4 times a day and operates on daily timeframes. Below are backtest results after the first calibration.

Win rate around 56-57% on 7d
For BTC this sounds realistic. If the terminal showed 70% over a long history from 2018-2026, I would say there is almost certainly data leakage or overfitting.

• BULL baseline WR7d = 56.7%

• BEAR baseline WR7d = 56.2%
  This looks like a real market where there is a positive edge, not magic.

2. Average 7d return also without fantasy

• BULL baseline Avg7d = +2.00%

• BEAR baseline Avg7d = +1.59% directional
  For BTC this looks plausible: volatility is high, so +1.5%...+2.0% over 7 days as conditional expectancy is normal. Not too little and not too good, and we are talking about the spot market, without leverage, where we consistently get a stable result.

3. Best regimes are not too perfect

• BULL YES Avg7d = +2.87%

• BEAR YES Avg7d = +3.80%
  This is already good, but still within market realism for BTC. Especially on a short horizon, where a strong regime filter can indeed pull such numbers.

How to get access?

Access will be opened for all paid subscribers of Adler Insight 💎 Premium. The terminal will work strictly through an active Substack subscription - no separate registration will be required as such. Access mechanics: upon login you enter the email your subscription is registered to, and if the subscription is active, the system sends a password to your email. As long as the subscription remains active, access to the terminal is maintained. The target timeline for the stable version release is currently April.

That’s all for news, let’s move directly to assembling all key signals of the week within the updated Weekly Engine. What the market has already confirmed. Where vulnerabilities remain. And what final signal the system gives for the next week. 👇