MPI + HMM Regime-Conditional Composites

MPI + HMM Regime-Conditional Composites

The MPI is a 0–100 composite of market positioning. The HMM gives you a regime classification with confidence. Used in isolation, each tells half the story. Combined into a regime-conditional composite, the two collapse into a single actionable decision frame that handles every realistic market state.

8.5.1Two Lenses Onto the Same Decision

The Market Pulse Index (MPI) and the regime model answer different questions about the same market state. The MPI answers “how favorable is the current configuration of positioning, breadth, and flow?” on a smooth 0–100 scale. The regime model answers “which of three regimes is the market in?” with a probability vector. The MPI is continuous and rich; the regime is discrete and structural. Either alone is incomplete; together they form a 2-D state space in which every realistic market condition has a defined response.

The most common misuse is to read one and ignore the other. Traders who read only the MPI miss the regime context that gives the score meaning — an MPI of 60 means something completely different in a Bull regime versus a Crisis regime. Traders who read only the regime miss the within-regime gradation — a Bull regime with MPI of 78 is a different trade environment from a Bull regime with MPI of 52. The composite framework forces both to be read together.

The output of the composite is a single decision frame for the day or week. The frame answers three questions: what is the maximum risk per trade, what is the new-entry quality bar, and what is the hedge ratio? All three answers depend on the joint state of MPI and regime, not on either alone. The discipline is to never act on one without the other.

The MPI tells you how good the configuration is. The regime tells you which game is being played. Both are required to know how to play.

8.5.2The Three Quadrant Rules

The composite collapses the joint MPI-regime state into three primary quadrants that cover the vast majority of realistic conditions. The remaining configurations are edge cases handled below.

Quadrant 1: MPI > 65 + Bull regime = Full Risk-On. Both lenses agree: the configuration is favorable and the regime supports it. New entries are taken at full regime-tier sizing on regime-matched setups. The hedge ratio is at baseline. Conviction-grade ideas can be added with reasonable confidence that the framework is in alignment. This is the quadrant in which the framework is doing its best work.

Quadrant 2: MPI 50–65 + Any Regime = Wait for Confirmation. The configuration is mid-pack and the framework is not strongly committed. New entries require both quality and corroboration; mediocre setups are passed. Sizing is at half the regime tier. The hedge ratio is at baseline-plus-25%. The discipline is patience: the framework is telling you it does not have a strong view, and the trade is to wait for clearer signal rather than to manufacture one.

Quadrant 3: MPI < 50 + Crisis regime = Defensive. Both lenses agree: configuration is poor and the regime is hostile. New entries are limited to the highest-quality setups, sized at quarter-tier. The hedge ratio is lifted meaningfully above baseline. Existing positions are reviewed for stop adjustment. This is the quadrant in which framework discipline matters most because the temptation to revenge-trade or to bottom-fish is at its highest.

8.5.3Regime-Conditional MPI — Same Number Reads Differently

The deepest insight of the composite is that an MPI of 60 reads completely differently in different regimes. In a Bull regime, MPI 60 is a sub-par configuration — the framework expects MPI 70+ in a healthy Bull, so a 60 reading is the framework noting that something has weakened the underlying positioning even though the regime label has not flipped. The trade implication is to size lighter even within Bull, because the framework’s two lenses are diverging.

In a Crisis regime, the same MPI of 60 reads as bullish recovery. The Crisis regime expects MPI in the 25–45 band; an MPI of 60 in Crisis means the configuration has improved meaningfully ahead of the regime label. Historically, MPI rising into the 55–65 band during a Crisis regime precedes the regime label flipping back to Neutral or Bull within 5 to 15 trading days. The 60 in Crisis is a leading indicator of regime recovery in a way that 60 in Bull never could be.

In a Neutral regime, MPI 60 is mid-range and aligned: the framework’s two lenses agree the configuration is mid-pack. Sizing is at half tier and the trade is to be patient until either MPI rises (configuration confirms) or the regime classification firms (regime confirms).

The discipline is the framing: the MPI by itself is a number, not a verdict. The same number across different regimes is different evidence. Reading the MPI without the regime overlay is reading half the data, and the half you are missing is exactly the half that determines the meaning of the half you have.

8.5.4Today’s Composite Read

The current composite illustrates Quadrant 1 cleanly. The MPI today is 78, well above the 65 threshold. The regime posterior is 0.89 Bull, above the 0.75 actionability threshold. The two lenses agree: the configuration is favorable and the regime supports it. SPY is at $723.71 and the cross-asset composite (rates, FX, credit) is at +2 corroboration.

The implications are concrete. Per-trade sizing is at the full Bull tier (1.0% of account, posterior-weighted to 0.89%). Maximum gross heat is 5% across positions. New-entry quality bar is “regime-matched and meets standard quality criteria” — not the more selective bar that applies in Quadrant 2. Hedge ratio is at baseline (the framework default light hedge for tail protection, not the elevated hedge of Quadrant 2 or 3).

The watch list overlay is to monitor for either the MPI dropping below 65 (would shift to Quadrant 2 even with regime intact) or the regime posterior dropping below 0.75 (would shift to Quadrant 2 even with MPI intact). Either change moves the framework’s posture meaningfully without requiring both to shift. The composite is sensitive to weakness on either axis, which is why neither axis can be ignored in daily routine.

8.5.5Building the Composite into Daily Routine

The composite is most useful when it is read every morning before the open as a pre-market discipline. The procedure: pull the latest MPI score, pull the latest regime posterior, locate the joint state on the quadrant table, read the prescribed sizing tier and hedge ratio, and compare to your current book. If the book is not consistent with the prescribed posture, the morning task is to bring it into alignment before taking new trades.

The other place the composite belongs is in the trade journal. Every trade entry should be tagged with the composite quadrant at entry. Over time, the journal will reveal which quadrants you trade well and which you trade poorly. Most traders are best at Q1 (Full Risk-On) where the framework is doing the work for them; most are worst at Q2 (Wait) because patience is harder than action. Knowing your own performance by quadrant lets you adapt your participation rate accordingly.

The composite is also the basis for after-the-fact attribution. A losing week in Q1 is a different diagnosis from a losing week in Q3. The Q1 loss is most likely about individual trade selection or stop placement; the Q3 loss may be about whether the framework should have engaged at all. Tagging by composite makes this attribution mechanical rather than narrative.

8.5.6Common Mistakes

• Reading the MPI without the regime overlay. The same number means different things across regimes.
• Reading the regime without the MPI overlay. Within-regime configuration matters for sizing decisions.
• Treating the quadrant table as binary. The edges (Bull + low MPI, Crisis + high MPI) are the most actionable cases.
• Ignoring the regime-recovery signal. MPI rising into 55–65 in Crisis precedes regime flips by 5–15 days.
• Failing to tag trades with composite at entry. Without tagging, after-the-fact attribution is impossible.
• Trading the same way in Q2 as in Q1. Patience is the trade in Q2; new entries are not.