PARTIAL · MM-only
Hypothesis
The microprice — mid adjusted by book imbalance — is a better short-horizon fair-value estimate than the mid, improving passive fill timing.
Math — imbalance-weighted microprice
Weight the two sides by the opposite queue size:
$$ P_{\text{micro}} = \frac{Q_b\,P_a + Q_a\,P_b}{Q_a + Q_b} $$
Method
Compute microprice from L1, test as next-tick mid predictor and as a passive-quote anchor.
Results
| Next-tick mid prediction | better than mid |
| Taker exploitation after fees | none |
The microprice genuinely predicts the next mid better than the mid — but the gain is sub-fee and only useful to a passive market maker choosing which side to rest on. No taker edge. Filed with the MM research (cf. N-026).
Better fair-value estimates help whoever is already quoting. For a taker paying the spread, a 0.3-tick improvement is invisible.