Solutions · Capital markets

Know what your signal is really trading on.

Attribute every forecast to its inputs, see which regime the model learned in, and catch decay while it is still a line on a chart. For quantitative research, risk and model validation teams.

Data
49 industriesmonthly US industry returns to 2026; 54,279 industry-months
In sample, then after
55.2% → 50.3%hit rate to 2009, and over the ten years that followed
What it trades on
39.5%of measured importance sits in one input, momentum rank
Fragility
3.7%of calls flip under 5%-σ input noise; 99% of forecasts sit within 0.10 of a coin flip

Where it fits

For signals, risk models and the people who sign them off.

Any model that makes one forecast from a fixed set of features: a probability, a return, a volatility, a cost.

Research

Signal attribution

Why the model favours one name over another today, in the units of the forecast: probability, basis points or expected return.

Risk

Risk and pricing models

Volatility, liquidity, prepayment and credit-spread models are numeric predictions. Kaimei explains each one in its own units and tracks it for drift.

Validation

Independent model review

Behaviour curves show whether a model responds to an input the way theory says it should, and whether that response is monotonic.

Production

Decay and regime monitoring

Every input compared with the training distribution on a schedule. Attach realised outcomes and track hit rate or error by day, week or month.

Release

Research to production

Ship a retrain as a new version and compare its insights with the one it replaces before it takes risk.

Execution

Cost and execution models

Slippage and fill models built on fixed features, explained fill by fill.

A worked example

A sector signal, opened up.

Public data49 Industry Portfolios, Kenneth R. French Data Library (Fama & French).sector-rotationscikit-learn · gradient boosting

A gradient-boosted model estimates the chance that an industry beats the median industry next month, from nine inputs known at month end: momentum, reversal, volatility, beta and three market-wide measures. It is trained on July 1963 to December 2009 and scored on every month since. A research example, gross of costs, and not investment advice.

sector-rotation · Explain
scikit-learn · gradient boosting
raiseslowers
49 Industry Portfolios POST https://api.kaimei.ai/v1/explanations

Outcomes

The backtest was the best it ever looked.

Kaimei does not judge a strategy. It shows what the model does, and the picture here is plain: an edge of 5.2 points in sample that is close to nothing afterwards.

The attribution says why. In August 2026 the model's top pick is electronic equipment at 56.3%, and +5.9 of the 7.2 points above average come from a single input: its momentum rank. A model that is mostly one well-known factor inherits that factor's fate.

Hit rate by year · sector-rotation1990 to 2026
'90
'95
'00
'05
'10
'15
'20
'25
50%
blue: in the training window (to 2009)pink: after it
In sample the model is right 55.2% of the time. In the ten years after training, 50.3%; since 2020, 51.8%. The top-minus-bottom quintile spread falls from 1.58% a month (t = 9.3) to 0.39% (t = 1.3).

What it learned

One factor, a thin margin.

momentum_rank is the most influential input, followed by reversal_1m and market_volatility_12m. Together they carry 68% of the measured importance (permutation importance over 2000 rows).

3.7% of decisions flip under 5%-σ input noise. Most fragile inputs: momentum_rank, volatility_12m. Kaimei rates that medium severity and raises it as an alert: On 2,000 reference rows the model predicts outperform 42% of the time; 99% of cases sit within ±0.10 of the decision threshold.

  • Higher momentum_rank increases the probability of outperform, most sharply between 0.8425 and 0.9309 (range 46% → 54%).
  • Higher reversal_1m increases the probability of outperform, most sharply between -3.314 % and -1.709 % (range 47% → 50%).
Global importance · sector-rotationpermutation · 2,000 rows
Momentum rank39.5%
Latest month18.9%
Market volatility10.1%
Market beta9.1%
Market, 12 months8.1%
Momentum, 12–1 months5.3%
Volatility, 12 months4.8%
Industry dispersion3.0%
Return, 36–13 months1.3%
momentum_rank is the most influential input, followed by reversal_1m and market_volatility_12m. Together they carry 68% of the measured importance (permutation importance over 2000 rows).

Regime monitoring

The market it trades is not the one it learned.

Each five-year window of live forecasts is compared with the 1963 to 2009 training rows. By 2015 to 2019 the market-wide inputs had already left the training distribution, and 2020 to 2024 adds the industries' own volatility. Stock-level inputs such as momentum rank stay stable throughout: the signal looks the same, and the conditions around it do not.

2015 to 2019significant
Market, 12 months1.78
Market volatility0.93
Industry dispersion0.23
Momentum, 12–1 months0.20
Volatility, 12 months0.20
Return, 36–13 months0.08
Recent traffic has significant drift (PSI 1.78). Shifted inputs: market_return_12m, market_volatility_12m, dispersion_1m, momentum_12_1. 2,940 logged predictions against 5,000 reference rows.
2020 to 2024significant
Industry dispersion1.89
Market volatility1.86
Volatility, 12 months0.37
Market, 12 months0.34
Latest month0.05
Market beta0.04
Recent traffic has significant drift (PSI 1.89). Shifted inputs: dispersion_1m, market_volatility_12m, volatility_12m, market_return_12m. 2,940 logged predictions against 5,000 reference rows.

Population stability index per input. Bars are cut off at 0.5; the line marks 0.25, where Kaimei calls drift significant.

One forecast

March 2020, explained after the fact.

At the end of March 2020 the model ranks aircraft last of 49, at 43.9%. The reasons are on the record: a −35.8% month (−2.5 pts), weak momentum (−2.6 pts) and a beta of 1.61 (−1.6 pts).

The industry returned +7.0% the next month against a median of +12.8%. Any logged forecast can be explained this way later, by its prediction id.

Top and bottom of the latest rankingAugust 2026 · P(outperform)
Chips56.3%
LabEq55.5%
Mines54.3%
Mach53.5%
Drugs53.2%
BusSv45.4%
Fun45.4%
PerSv45.4%
Autos44.6%
Clths42.4%
French's industry abbreviations. Five highest and five lowest of 49.

Who it is for

Research, risk and the desk, reading the same record.

01

Quantitative research

See which inputs a model really uses before it is sized. Compare two versions of a signal by what they learned, not only by their backtests.

02

Risk and model validation

Check direction and monotonicity against theory, measure fragility under input noise, and keep the evidence with the model's record.

03

Portfolio management and trading

Ask why for any position the model drives, in the dashboard, and see an alert when the inputs leave the range the model was trained in.

Rules and guidance

Evidence for model and algorithm governance.

Rule or guidanceWhat it asksWhere Kaimei helps
Algorithmic trading in the EUMiFID II, Commission Delegated Regulation (EU) 2017/589 (RTS 6)Testing of algorithms before deployment, real-time monitoring and an annual self-assessment.A stored analysis at each release, drift checks on a schedule, and a record of every forecast and its explanation.
Algorithmic trading in the UKPRA Supervisory Statement SS5/18Governance, approval, testing and an inventory of algorithms.Versioned models with their insights, exportable for the inventory.
Model risk guidanceUS interagency guidance of April 2026 (SR 26-2); PRA SS1/23 in the UKValidation and ongoing monitoring of pricing, risk and valuation models.Behaviour curves, sensitivity and stability evidence, and monitoring with alerts.

Coming soon

Next on the roadmap.

What we are building now, and what to use until each one ships.

  • Coming soonSequence models

    Models that read ticks or a price history step by step, with each prediction attributed to time steps as well as inputs. Until then, a model on lags, rolling statistics and other fixed features works.

  • Coming soonSeveral outputs per model

    Every horizon or target of one model explained in a single call. Until then, register one model per forecast horizon or target.

  • Coming soonAttribution drift

    An alert when what the model relies on shifts, not only its inputs: a signal that stops mattering, or one that starts to carry the model.

  • Coming soonChallenger benchmarks

    A simple, interpretable model fitted to the same reference data, so validators can see what the complex model adds. Until then, register a challenger yourself and compare the two side by side.

See it on your own model.

Bring a signal or a risk model. We will show you what it is leaning on.