Predictive Signal Analysis

Predictive Signal Scarcity Within Diversified Financial Services Equity

Current data demonstrates zero predictive price signals and insufficient institutional flow correlation for modeling

MS • 2026-03-04

12A: Price Signals vs Fundamental Outcomes

How to read this section: We test whether three price-based signals — 12-month momentum (trailing stock return), realized volatility (annualized standard deviation of daily returns), and relative strength (stock return minus S&P 500 return) — predict next-quarter fundamental outcomes: revenue growth, operating margin change, and ROE change (all year-over-year to remove seasonality). Each cell shows the Pearson correlation (r) between signal at quarter Q and outcome at quarter Q+1. Values closer to +1 or −1 indicate stronger predictive relationships. “n” is the number of paired observations.

Analysis of Morgan Stanley (MS) from 2015Q1 to 2025Q4 reveals a total absence of statistically significant predictive relationships between price-based signals and fundamental outcomes. Across 40 observation quarters, traditional indicators such as 12M momentum, realized volatility, and relative strength failed to provide a reliable lead for revenue growth, margin expansion, or ROE shifts. For an institutional investment bank and wealth manager like MS, the lack of signal suggests that price action is likely a coincident reflection of macro-environmental factors—such as interest rate cycles and capital market volumes—rather than a lead indicator of idiosyncratic fundamental shifts.

Morgan Stanley (MS) 44 quarters | 2015Q1 to 2025Q4
Signal \ Outcome Revenue Growth Margin Change ROE Change
12M Momentum 0.11
n=40
weak
0.01
n=40
weak
0.01
n=40
weak
Realized Volatility 0.10
n=40
weak
-0.21
n=40
weak
0.19
n=40
weak
Relative Strength -0.08
n=40
weak
-0.19
n=40
weak
-0.09
n=40
weak
Strongest: No notable signals found

Morgan Stanley exhibits no notable predictive signals within the studied parameters. The strongest observed relationship, realized volatility versus margin change, yielded a correlation of r=-0.211 (n=40, p=0.192), which is statistically insignificant and explains less than 5% of the variance. Furthermore, 12M momentum shows effectively zero predictive power for ROE changes (r=0.013, p=0.938), indicating that historical price trends do not anticipate shifts in the firm's capital efficiency. The results suggest that MS equity pricing is efficiently integrated with fundamental disclosures or driven by exogenous variables not captured by these specific price-to-fundamental lags.

MS - Correlation Heatmap

Cross-Company Patterns

No consistent cross-company predictive signals found

12B: Institutional Flow vs Price Impact

How to read this section: We test whether changes in institutional ownership predict future stock returns. Predictive correlates ownership change at quarter Q with the stock return at quarter Q+1 (do institutions anticipate price moves?). Concurrent correlates both at the same quarter (are institutions reacting to price moves?). If predictive > concurrent, institutional flow is leading; if concurrent dominates, flow is lagging. Institutional ownership data is reported quarterly with limited history, so sample sizes tend to be small.

Analysis of institutional ownership changes for Morgan Stanley (MS) indicates a predominantly concurrent relationship between capital flows and price performance. The data demonstrates that institutional positioning shifts are synchronized with price moves rather than preceding them, as evidenced by a notable concurrent correlation (r=0.56, n=6) and a negligible predictive correlation (r=-0.14, n=5). This pattern suggests that institutional activity in MS is largely reactive, likely driven by momentum-following strategies or rebalancing in response to realized price volatility.

Morgan Stanley (MS) concurrent
Metric Correlation p-value n Significance
Predictive (flow Q → return Q+1) -0.1432 0.8184 5 weak
Concurrent (flow Q ↔ return Q) 0.5565 0.2514 6 notable
Concurrent |r|=0.56 exceeds predictive |r|=0.14 by >0.1

Morgan Stanley is classified as showing a concurrent institutional flow pattern. The concurrent correlation of r=0.5565 (n=6, p=0.2514) is considered notable, suggesting that institutional net buying or selling explains approximately 31% of the variance in price within the same quarter. However, the predictive signal is weak (r=-0.1432, n=5, p=0.8184), indicating that prior-quarter institutional flows have no statistically significant relationship with subsequent price action. The disparity between concurrent and predictive values (|r| difference of 0.42) confirms that institutions are following price trends rather than leading them.

MS - Ownership Change vs Next-Quarter Return

12C: Earnings Surprise Patterns

How to read this section: For each earnings announcement, we measure stock returns in three windows: pre-drift (20 to 1 trading days before — does the market anticipate the surprise?), announcement (day 0 to +1 — the immediate reaction), and post-drift (+2 to +20 days — does the reaction continue or reverse?). Events are classified as positive (>2% EPS surprise), negative (<−2%), or inline. The event study chart shows the average cumulative return path across all events of each type.

Morgan Stanley (MS) exhibits a consistent pattern of positive earnings surprises across the observation period (n=4), maintaining a 100% beat rate for both EPS and revenue. The magnitude of EPS surprises is significant, averaging 22.06%, while revenue surprises are comparatively modest at 1.82%. This divergence suggests that bottom-line outperformance is driven more by margin expansion or capital markets activity than by top-line growth alone. The return profile is characterized by positive drift in all three phases: pre-announcement (+3.30%), announcement reaction (+3.18%), and post-announcement drift (+1.22%). A notable statistical feature is the strong inverse relationship (r=-0.6312, n=4) between pre-announcement drift and the subsequent surprise. This strong negative correlation suggests that as the market bids up the stock prior to the event, the realized surprise often fails to catalyze further outsized gains relative to expectations, or conversely, that periods of lower pre-event momentum precede the largest positive surprises. This relationship explains approximately 40% of the variance in surprise outcomes within this limited sample.

Morgan Stanley (MS) 4 events
Beat Rate
100.0%
Avg EPS Surprise
22.06%
Consecutive Beats
4
Surprise Trend
stable
Direction Events Avg Pre-drift [-20,-1] Avg Announcement [0,+1] Avg Post-drift [+2,+20]
positive 4 3.30% 3.18% 1.22%
Pre-drift return predicts surprise direction (r=-0.6312)

Morgan Stanley demonstrates high consistency with four consecutive beats and a stable surprise trend. The average announcement effect of 3.18% indicates that despite the 100% beat rate, the market continues to underprice the magnitude of MS's earnings strength. The strong negative correlation (r=-0.63) between pre-drift and the surprise direction suggests that the pre-announcement price action acts as a contrarian indicator for the surprise magnitude rather than reflecting information leakage. The post-announcement drift of 1.22% suggests a moderate 'post-earnings announcement drift' (PEAD) effect, where the market takes additional time to fully integrate the new fundamental data.

MS - Earnings Event Study [-20, +20] Days

12D: Multi-Signal Integration

The quantitative assessment of Morgan Stanley (MS) reveals a significant divergence between continuous fundamental-price coupling and discrete event-driven predictability. While traditional price-fundamental signals fail to reach the threshold for notable correlation (|r| < 0.40), the equity demonstrates high predictability within earnings-event regimes. This suggests that the market's mechanism for pricing MS is heavily weighted toward quarterly reporting cycles rather than ongoing fundamental discovery.

Company Price-Fundamental Signals Institutional Predictive Pre-drift Predictive Earnings Consistency Signal Coverage Data Quality
MS 0 No Yes consistent beater moderate strong
MS

Morgan Stanley (MS) exhibits notable pre-drift predictive power, which converges with a 100% earnings beat rate to suggest a highly patterned response to quarterly results. Despite this, the stock shows an absence of notable or strong price-fundamental signals (0 signals with |r| >= 0.40), indicating that valuation metrics do not reliably lead price trends over the observed period. Institutional predictive signals are non-existent, suggesting that order flow data lacks a leading relationship with subsequent returns. While signal coverage is moderate, the strong data quality provides high confidence in the observed earnings consistency, identifying MS as a 'patterned' asset primarily during reporting windows rather than a trend-following fundamental play.

Signal Coverage Heatmap

12E: Signal Discovery Summary

Signal discovery analysis for Morgan Stanley (MS) identifies a strong inverse relationship between pre-announcement price drift and subsequent earnings surprises (r = -0.6312, n=4). This suggests that price action in the 20 trading days leading up to an announcement has historically acted as a contrarian indicator for the magnitude of fundamental surprises. Specifically, the negative correlation indicates that periods of price weakness or consolidation have preceded positive earnings surprises, while recent price strength has correlated with smaller surprises or disappointments. Beyond price-fundamental relationships, MS has demonstrated a consistent streak of four consecutive earnings beats, suggesting a persistent gap between consensus analyst expectations and realized performance. However, the analysis failed to identify any consistent predictive signals that persist across multiple companies, indicating that market signals in this sector remain highly idiosyncratic and resistant to broad-based heuristic modeling. While the identified signal for MS is statistically strong (|r| > 0.6), the small sample size of four earnings events limits the reliability of the finding. The observed relationship may be regime-dependent and should be treated as a preliminary observation rather than a robust predictive law. The lack of cross-company patterns further emphasizes the need for security-specific analysis over generalized sector signals.

Signal Predictability Rankings

MS moderate

Strong negative correlation (r = -0.6312) between 20-day pre-earnings price drift and earnings surprises suggests a contrarian signal pattern.

Top Signals by Company

Morgan Stanley (MS)
Pre-drift return predicts earnings surprise: r=-0.6312
4 consecutive earnings beats
Caveats: Correlation does not imply causation; Past predictive relationships may not persist

Monitoring Recommendations

Monitor MS price drift in the [-20, -1] day window prior to earnings for contrarian signaling
Track the persistence of the 4-quarter earnings beat streak as an indicator of analyst model lag
Evaluate shifts in the r = -0.6312 correlation coefficient as new quarterly data points are added to the sample

Cross-Company Patterns

No consistent cross-company predictive signals found

Key Takeaways

1. 1. Morgan Stanley exhibits a strong negative correlation (r = -0.6312) between pre-drift returns and earnings surprises.
2. 2. A streak of 4 consecutive earnings beats indicates a sustained period of conservative consensus estimates for MS.
3. 3. No universal predictive signals were identified across the broader cross-asset sample, highlighting the uniqueness of MS's price-fundamental relationship.
4. 4. The high correlation coefficient in MS is mitigated by a small sample size (n=4), necessitating cautious application.

Methodology

Signal discovery uses Pearson correlation with lagged variables. Minimum sample sizes: 8 quarterly observations for price-fundamental, 5 for institutional flow, 4 earnings events. Significance thresholds: |r| >= 0.6 (strong), |r| >= 0.4 (notable). All correlations are bivariate; multivariate relationships not tested. Quarterly fundamentals use YoY changes (pct_change(4)) to avoid seasonality. Event study uses trading days [-20, +20] around earnings announcements.
Analysis utilizes bivariate Pearson correlations with lagged variables. Minimum sample sizes of 4 earnings events and 8 quarterly observations apply. These findings identify statistical relationships only; they do not imply causation and are subject to significant regime-dependence and small-sample bias.

Related Reports

Risk Alert Panel
Structural Capital Erosion and Heightened Risk at Morgan Stanley
MS
Institutional & Insider Activity
Institutional Demand Offsets Notable Insider Liquidation Patterns
MS
Returns & Risk Profile
Wealth Management Resilience Balanced Against Capital Markets Sensitivity
MS