Finance

Neural Generated Portfolios: The paper proposes a new neural-network-based method for learning the generating function of a functionally generated portfolio from market data, offering a strong alternative to traditional portfolio optimization. (2025-06-24, shares: 3)

Dependence Measures for Portfolios: The paper introduces a new method, NAbC,…

Read more

Read more

At Sov.ai Research we empower investors with rich data and analytics. Our offerings include insightful newsletters and an easy-to-use Python SDK. Enjoy this week’s free newsletter! These signals were already available by Sunday in the data product.

If you are interested in working with us to develop an internal agentic finance project, get in touch at, we are happy to share testimonials:

AgenticFinance.ai

• Financial Health:

  • PEGA demonstrates strong financial stability with high Cash Flow (CF: 13) and Liquidity (LIQ: 41) scores, positioning it for medium-term growth despite recent stock performance.

  • GLPG exhibits concerning financial metrics, particularly in Profitability (-48) and Solvency (25), indicating deteriorating health compared to industry peers.

• Insider Flow Pressure:

  • HD and ESQ present promising buying opportunities in the Consumer Cyclical and Financial Services sectors, respectively, with ESQ showing a notable flow prediction change of 25.71.

  • PIII and CAT show significant selling pressures, with CAT demonstrating a concerning flow prediction change of -43.41 in the Industrials sector.

• Short Interest Analysis:

  • PIII is heavily overshorted (+17.8% change) at 88.5% vs expected 39.2%, making it a potential short squeeze candidate.

  • VRM's declining short interest (-13.8% change) suggests bearish trends, while SANG and CUB indicate shifting short volumes that could impact market movements.

• Earnings Surprise Prediction:

  • BOLD is likely to exceed earnings expectations with a 50.8% likelihood of positive surprise, whereas SAR faces substantial downside risk with a -50.9% probability.

• Breakout Prediction:

  • K shows strong potential for upward price movement, while ALXO and CWCO indicate possible downward pressures.

• Sentiment Modelling:

  • BVFL experiences the most positive news sentiment with a 62.151 change in score, contrasted by SASR, which faces significant negative media coverage with a -65.306 change.

This information is for general purposes only and does not constitute financial, investment, tax, or legal advice. Nothing here should be taken as a recommendation to buy or sell any securities or financial instruments. All content is provided "as is" without warranties, and may not suit your personal circumstances or financial objectives.

Financial Ratio Radar

This is a financial early-warning system that monitors changes in ratios to identify critical positive and negative trends in companies' relative performance. It focuses on 8 key areas:

• Composite (Comp): Combines all 6 measures into one number

• Cash Flow (CF): Change in how well the company manages its cash

• Efficiency (EF): Change in how well they use their resources

• Liquidity (LIQ): Change in how easily they can pay short-term bills

• Profitability (Prof): Change in how much money they're making

• Solvency (Solv): Change in how well they can handle their debts

• Valuation (Value): Change in how fair their stock price is

• Factor (Fac): A measure of statistical change (can ignore)

  YOU CAN SORT THE COLUMNS

We supply 650+ of the the largest positive and negative moving ticker ratios from the week.

• PEGA demonstrated strong financial health with high scores across multiple metrics, particularly in CF (13) and Liquidity (41), outperforming peers in overall financial stability. However, VRT looks like a better long target considering its lack of price movement.

• GLPG showed concerning financial metrics, with notably weak scores in Profitability (-48) and Solvency (25), indicating deteriorating financial health compared to industry peers. However, AMBP and TPH looks like better short targets, considering their inverse price movements.

• Key Trends to Watch:

  • Positive: Companies improving in both Cash Flow and Profits (top right of the plot) often make great investments—they’re getting better at making and retaining money. VRT is showing up once again.

  • Negative: Companies struggling with both Cash Flow and Profits (bottom left) indicate potential issues.

Insider Flow Pressure

This model forecasts insider trading behavior, the idea is to predict when insiders will buy and sell stocks and to ‘front-run’ them. This model self-improves over time using machine learning. The tables focus on the top 20 stocks with the highest buying pressure and the top 20 stocks with the highest selling pressure.

Shows how a stock's predicted pressure compares to similar companies, ranked as a percentage

• HD shows the most promising flow prediction indicating a potential buying opportunity within the Consumer Cyclical sector.

• PIII has the most concerning flow prediction suggesting significant selling pressure in the Healthcare sector.

Looking at the change in a prediction from one week to another gives you a second-order output that is often a better leading indicator to stock price movements.

• ESQ shows a promising change in flow prediction of 25.71, indicating a potential buying opportunity within the Financial Services sector.

• CAT shows a concerning change in flow prediction of -43.41, suggesting significant selling pressure in the Industrials sector.

Short Interest Analysis

We identify stocks whose short interest significantly deviates from comparable companies, refreshed every two weeks. The table focuses on the top and bottom 60 overshorted stocks. Red is undershorted and Blue is overshorted

• PIII is heavily overshorted (+17.8% change) with 88.5% vs expected 39.2% - potential squeeze candidate. Very interesting that it was twice mentioned (institutional, and overshorted above)

• VRM shows declining short interest (-13.8% change) at 11.0% vs expected 61.1% - this is bearish given that future shorts could catch on.

By the time the short-interest file becomes available, the market may have already reacted. Monitoring changes in short volume, however, could provide an earlier indication of shifts in short sentiment. Displays top and bottom 150 stocks by short volume ratio changes.

• SANG shows rising short volume (+82.8% change) to 99.2% vs previous 16.4% and is a potentially concerning trend.

• CUB has declining short volume (-60.2% change) to 2.1% vs previous 62.3% and could signal covering with upside potential.

Earnings Surprise Prediction

Predicts earnings surprise for companies reporting in the next two weeks. By anticipating unexpected earnings results, you can position yourself ahead of market movements.

• The table only shows company that are reporting in the next two weeks from the date indicated at the top of the table.

• The earnings surprise predictions are bounded between -50% and +50% to filter out statistical outliers and maintain focus on meaningful variations.

• BOLD is predicted to potentially outperform expectations with a 50.8% likelihood of earnings surprise. The consensus EPS estimate is $-0.68 for the quarter ending 2025-03-30.

• SAR shows potential downside risk with a -50.9% earnings surprise probability. Analysts expect EPS of $0.77 for the quarter ending 2025-03-01.

Breakout Prediction

Using machine learning models, we track the predicted change in price breakouts and slope. The implication of a change is that a stock has become more or less likely to go up in price. We deliver the 80 tickers with the largest and smallest predicted changes in breakout and slope.

• K shows the strongest breakout prediction signal, suggesting potential upward movement. Looks like a M&A false positive.

• ALXO displays the weakest breakout prediction signal, indicating possible downward pressure.

• GEN demonstrates the largest positive change in breakout prediction, suggesting improving momentum.

• CWCO shows the largest negative change in breakout prediction, indicating deteriorating momentum.

• AFBI exhibits the steepest positive change in breakout slope, suggesting accelerating upward momentum.

• GBDC shows the steepest negative change in breakout slope, indicating accelerating downward pressure.

Sentiment Modelling

Tracking news-based sentiment for individual companies allows us to aggregate sentiment to the sectorial level. We first deliver the 280 tickers with the largest and smallest predicted changes in sentiment. The aggregated data is then presented on a sectorial and thematic basis.

• News sentiment for BVFL articles was most positive this week with a 62.151 change in sentiment score, suggesting favorable coverage.

• Meanwhile, SASR faced the most negative news sentiment shift of -65.306, indicating negative media coverage.

The chart above tracks the sentiment at a sectorial level which is helpful to assess weekly sectorial sentiment shifts. An increasing percentile indicates improving sentiment for a given sector/theme relative to others.

We also track numerous themes that we aggreagate in a handful of main themes, the detailed themes are available in the SDK. An increasing percentile indicates improving sentiment for a given sector/theme relative to others.

Congressional Trading Tracker

This tool tracks filings and records in both the House and Senate, based on the premise that members of Congress may be more informed than the average investor. The table shows trades performed in a given week.

Consumer Financial Complaints

We investigate recent changes in risk for financial firms based on complaints received in the latest week available. We deliver all the tickers that experienced a complaint this week with their change in risk. Please confirm the ticker-company name label, the mapping is still in beta and will improve over time.

The total risk rating is an ML-derived score measuring potential company liability based on complaint severity rather than volume.

• DHX received 5 consumer complaints in the preceding week, with a concerning rise in financial risk of 9.67 points and a notably high culpability score of 96.46.

• Despite 6 consumer complaints, CRMT shows significant improvement with financial risk dropping 12.82 points, while maintaining a customer grievance score of 19.08.

Government Spending

This data tracks recent government contract announcements. Given that the government accounts for over 40% of national expenditures, these contracts are an important source of revenue for many companies and are a reliable predictor of quarterly revenue. We deliver all the tickers with new government contracts for the associated week.

• Size indicates the size of the contract in USD millions. Obligation is the contractual amount the government ought to pay the company.

• Awards corresponds to the number of contracts awarded in the associated week. Previous indicates the number of contracts previously awarded to the same company.

Corporate lobbying tracker

Lobbyists have important roles to play in the American political system and are often responsible for the passage of bills and implementation of policies.

Anomaly corresponds to a change in lobbying behavior compared to historical lobbying patterns. It could indicate that it is a companies first lobbying contract, or a large contract compared to their historical average.

Wikipedia Search Interest

This model uses an algorithm to discover early search interest on Wikipedia pages. This signal is stronger than most retail indicators as it suggests the beginning of a research process. We deliver the 100 tickers with the largest and smallest predicted changes in search behaviour.

• Interest in IMBI, based on Wikipedia searches, increased 32.57% this week, showing strong positive attention.

• Meanwhile, RSKD experienced a significant decline of -26.80% in search interest, indicating reduced market attention.

Factor Model Coefficients

This analysis shows percentile rankings of factor model coefficients across stocks. Higher percentile coefficients indicate stronger factor sensitivity relative to peers. We deliver approximately 1000 tickers and their associated factor scores

Factor Model Error Analysis

Stocks with high R², negative AIC, and strong t-statistics provide reliable factor exposure for long-term systematic investing, while those with lower statistical significance and model fit offer alpha potential through market inefficiencies but demand sophisticated trading strategies.

Thank you for reading and subscribing. We aim to add a new screen every month and make it available to you on Mondays.

Key Insights

• Financial Health:

  • CVAC demonstrates strong financial stability with high Cash Flow (CF) and Liquidity (LIQ) scores, positioning it for medium-term growth despite recent stock declines.

  • PRTA exhibits concerning financial metrics, particularly in Profitability and Solvency, indicating deteriorating health despite buoyant stock performance.

• Insider Flow Pressure:

  • AMP and GM present promising buying opportunities in the Financial Services and Consumer Cyclical sectors, respectively.

  • NKLA and BZH show significant selling pressures in the Industrials and Consumer Cyclical sectors, signaling potential risks.

• Short Interest Analysis:

  • ARBB is heavily overshorted, making it a potential short squeeze candidate.

  • CYCC's declining short interest suggests bearish trends, while CCIX and FTS indicate shifting short volumes that could impact market movements.

• Earnings Surprise Prediction:

  • ZIM is likely to exceed earnings expectations, whereas ACM faces substantial downside risk.

• Breakout Prediction:

  • EVRI shows strong potential for upward price movement, while RILY and FFNW indicate possible downward pressures.

• Sentiment Modelling:

  • TT experiences the most positive news sentiment, contrasted by AIN, which faces significant negative media coverage.

Additional tools included in the report track congressional trading, government spending, corporate lobbying, Wikipedia search interest, and factor model analyses, providing a multifaceted view of market dynamics.

This information is for general purposes only and does not constitute financial, investment, tax, or legal advice. Nothing here should be taken as a recommendation to buy or sell any securities or financial instruments. All content is provided "as is" without warranties, and may not suit your personal circumstances or financial objectives.

Financial Ratio Radar

This is a financial early-warning system that monitors changes in ratios to identify critical positive and negative trends in companies' relative performance. It focuses on 8 key areas:

• Composite (Comp): Combines all 6 measures into one number

• Cash Flow (CF): Change in how well the company manages its cash

• Efficiency (EF): Change in how well they use their resources

• Liquidity (LIQ): Change in how easily they can pay short-term bills

• Profitability (Prof): Change in how much money they're making

• Solvency (Solv): Change in how well they can handle their debts

• Valuation (Value): Change in how fair their stock price is

• Factor (Fac): A measure of statistical change (can ignore)

We supply 650+ of the the largest positive and negative moving ticker ratios from the week.

• CVAC demonstrated strong financial health with high scores across multiple metrics, particularly in CF (49) and Liquidity (24), outperforming peers in overall financial stability. Despite the stock's decline, being mostly a result of its German and an mRNA association, it is well positioned for medium term advances.

• PRTA showed concerning financial metrics, with notably weak scores in Profitability (-50) and Solvency (-30), indicating deteriorating financial health compared to industry peers. The stock is suprisingly boyant for its level of deteriorating health.

• Key Trends to Watch:

  • Positive: Companies improving in both Cash Flow and Profits (top right of the plot) often make great investments—they’re getting better at making and retaining money. Keep your eye on SGMO!

  • Negative: Companies struggling with both Cash Flow and Profits (bottom left) indicate potential issues. Watch out for ATEX!

Insider Flow Pressure

This model forecasts insider trading behavior, the idea is to predict when insiders will buy and sell stocks and to ‘front-run’ them. This model self-improves over time using machine learning. The tables focus on the top 20 stocks with the highest buying pressure and the top 20 stocks with the highest selling pressure.

Shows how a stock's predicted pressure compares to similar companies, ranked as a percentage

• AMP shows the most promising flow prediction indicating a potential buying opportunity within the Financial Services sector.

• NKLA has appeared twice as the most concerning flow prediction suggesting significant selling pressure in the Industrials sector.

Looking at the change in a prediction from one week to another gives you a second-order output that is often a better leading indicator to stock price movements.

• GM shows a promising change in flow prediction of 31.39, indicating a potential buying opportunity within the Consumer Cyclical sector.

• BZH shows a concerning change in flow prediction of -46.85, suggesting significant selling pressure in the Consumer Cyclical sector.

Short Interest Analysis

We identify stocks whose short interest significantly deviates from comparable companies, refreshed every two weeks. The table focuses on the top and bottom 60 overshorted stocks. Red is undershorted and Blue is overshorted

• ARBB is heavily overshorted (+10.2% change) with 0.8% vs expected 45.5% - potential squeeze candidate.

• CYCC shows declining short interest (-24.7% change) at 20.0% vs expected 58.2% - this is bearish given that future shorts could catch on.

By the time the short-interest file becomes available, the market may have already reacted. Monitoring changes in short volume, however, could provide an earlier indication of shifts in short sentiment. Displays top and bottom 150 stocks by short volume ratio changes.

• CCIX shows rising short volume (+100.0% change) to 100.0% vs previous 0.0% and is a potentially concerning trend.

• FTS has declining short volume (-62.0% change) to 10.5% vs previous 72.5% and could signal covering with upside potential.

Earnings Surprise Prediction

Predicts earnings surprise for companies reporting in the next two weeks. By anticipating unexpected earnings results, you can position yourself ahead of market movements.

• The table only shows company that are reporting in the next two weeks from the date indicated at the top of the table.

• The earnings surprise predictions are bounded between -50% and +50% to filter out statistical outliers and maintain focus on meaningful variations.

• ZIM is predicted to potentially outperform expectations with a 46.3% likelihood of earnings surprise. The consensus EPS estimate is $7.05 for the quarter ending 2024-09-30.

• ACM shows potential downside risk with a -49.9% earnings surprise probability. Analysts expect EPS of $1.24 for the quarter ending 2024-09-30.

Breakout Prediction

Using machine learning models, we track the predicted change in price breakouts and slope. The implication of a change is that a stock has become more or less likely to go up in price. We deliver the 80 tickers with the largest and smallest predicted changes in breakout and slope.

• EVRI shows the strongest breakout prediction signal, suggesting potential upward movement.

• RILY displays the weakest breakout prediction signal, indicating possible downward pressure.

• LARK demonstrates the largest positive change in breakout prediction, suggesting improving momentum.

• FFNW shows the largest negative change in breakout prediction, indicating deteriorating momentum.

• EVRI exhibits the steepest positive change in breakout slope, suggesting accelerating upward momentum.

• JJSF shows the steepest negative change in breakout slope, indicating accelerating downward pressure.

Sentiment Modelling

Tracking news-based sentiment for individual companies allows us to aggregate sentiment to the sectorial level. We first deliver the 280 tickers with the largest and smallest predicted changes in sentiment. The aggregated data is then presented on a sectorial and thematic basis.

• News sentiment for TT articles was most positive this week with a 55.972 change in sentiment score, suggesting favorable coverage.

• Meanwhile, AIN faced the most negative news sentiment shift of -77.241, indicating negative media coverage.

The chart above tracks the sentiment at a sectorial level which is helpful to assess weekly sectorial sentiment shifts. An increasing percentile indicates improving sentiment for a given sector/theme relative to others.

We also track numerous themes that we aggreagate in a handful of main themes, the detailed themes are available in the SDK. An increasing percentile indicates improving sentiment for a given sector/theme relative to others.

Congressional Trading Tracker

This tool tracks filings and records in both the House and Senate, based on the premise that members of Congress may be more informed than the average investor. The table shows trades performed in a given week.

Government Spending

This data tracks recent government contract announcements. Given that the government accounts for over 40% of national expenditures, these contracts are an important source of revenue for many companies and are a reliable predictor of quarterly revenue. We deliver all the tickers with new government contracts for the associated week.

• Size indicates the size of the contract in USD millions. Obligation is the contractual amount the government ought to pay the company.

• Awards corresponds to the number of contracts awarded in the associated week. Previous indicates the number of contracts previously awarded to the same company.

Corporate lobbying tracker

Lobbyists have important roles to play in the American political system and are often responsible for the passage of bills and implementation of policies.

Anomaly corresponds to a change in lobbying behavior compared to historical lobbying patterns. It could indicate that it is a companies first lobbying contract, or a large contract compared to their historical average.

Wikipedia Search Interest

This model uses an algorithm to discover early search interest on Wikipedia pages. This signal is stronger than most retail indicators as it suggests the beginning of a research process. We deliver the 100 tickers with the largest and smallest predicted changes in search behaviour.

• Interest in EVRI, based on Wikipedia searches, increased 32.92% this week, showing strong positive attention.

• Meanwhile, CVD experienced a significant decline of -32.09% in search interest, indicating reduced market attention.

Factor Model Coefficients

This analysis shows percentile rankings of factor model coefficients across stocks. Higher percentile coefficients indicate stronger factor sensitivity relative to peers. We deliver approximately 1000 tickers and their associated factor scores

Factor Model Error Analysis

Stocks with high R², negative AIC, and strong t-statistics provide reliable factor exposure for long-term systematic investing, while those with lower statistical significance and model fit offer alpha potential through market inefficiencies but demand sophisticated trading strategies.

Thank you for reading and subscribing. We aim to add a new screen every month and make it available to you on Mondays.

The information provided is for general informational and educational purposes only and should not be construed as financial, investment, tax, or legal advice. Nothing contained herein constitutes a solicitation, recommendation, endorsement, or offer to buy or sell any securities or other financial instruments. All content and materials are presented without warranty of any kind, express or implied. The information provided does not take into account your personal circumstances, financial situation, or objectives and may not be suitable for your specific needs.

Executive Summary

This report provides an analysis of various companies based on financial ratios, insider trading predictions, short interest analysis, search interest, sentiment modeling, breakout predictions, consumer financial complaints, government spending, corporate lobbying, and earnings surprise predictions.

Financial Ratio Radar

The Financial Ratio Radar monitors changes in key financial metrics to identify critical trends in companies' performance.

• Zhibao Technology Inc. (ZBAO): Demonstrated strong financial health with high scores in Cash Flow (CF: 66) and Liquidity (LIQ: 12), outperforming peers in overall financial stability.

• Marqeta Inc. (MQ): Exhibited concerning financial metrics, with weak scores in Profitability (Prof: -25) and Solvency (Solv: -14), indicating deteriorating financial health compared to industry peers.

Insider Flow Pressure

This model forecasts insider trading behavior to predict potential buying or selling opportunities.

• Northern Trust Corporation (NTRS): Shows the most promising flow prediction, indicating a potential buying opportunity within the Financial Services sector.

• Nikola Corporation (NKLA): Has the most concerning flow prediction, suggesting significant selling pressure in the Industrials sector.

Short Interest Analysis

Identifies stocks with significant deviations in short interest compared to comparable companies.

• eLabDoc (ELAB): Heavily overshorted with a 12.5% increase, reaching 78.8% versus the expected 27.2%, indicating a potential short squeeze candidate.

• VSTE: Shows declining short interest with a 23.6% decrease to 3.3% versus the expected 67.4%, which is bearish as future shorts could catch on.

Wikipedia Search Pressure

Utilizes algorithms to detect early search interest on Wikipedia pages, indicating the beginning of a research process.

• Doximity Inc. (DOCS): Experienced a 32.04% increase in search interest, showing strong positive attention.

• TAL Education Group (TAL): Saw a significant decline of 27.57% in search interest, indicating reduced market attention.

Sentiment Modeling

Tracks news-based sentiment for individual companies and aggregates it at the sectorial level.

• Jabil Inc. (JBL): Received the most positive news sentiment this week with a 59.009 change in sentiment score, suggesting favorable coverage.

• Aclaris Therapeutics Inc. (ACRS): Faced the most negative news sentiment shift of -60.000, indicating challenging media coverage.

Breakout Prediction

Uses machine learning models to predict changes in price breakouts and momentum.

• Affinity Bancshares Inc. (AFBI): Shows the strongest breakout prediction signal, suggesting potential upward movement.

• Cabaletta Bio Inc. (CABA): Displays the weakest breakout prediction signal, indicating possible downward pressure.

Consumer Financial Complaints

Investigates recent changes in risk for financial firms based on consumer complaints.

• New York Community Bancorp Inc. (NYCB): Received 45 consumer complaints in the preceding month, with a concerning rise in financial risk of 6.21 points and a notably high culpability score of 88.55.

• SLM Corporation (SLM): Despite 40 consumer complaints, shows significant improvement with financial risk dropping 15.07 points, while maintaining a customer grievance score of 14.34.

Earnings Surprise Prediction

Predicts companies expected to experience an earnings surprise in the near future.

• Immix Biopharma Inc. (IMMX): Predicted to potentially outperform expectations with a 49.7% likelihood of earnings surprise. The consensus EPS estimate is $-0.16 for the quarter ending September 30, 2024.

• Live Nation Entertainment Inc. (LYV): Shows potential downside risk with a -50.5% earnings surprise probability. Analysts expect EPS of $1.60 for the quarter ending September 30, 2024.

Financial Ratio Radar

This is a financial early-warning system that monitors changes in ratios to identify critical positive and negative trends in companies' relative performance. It focuses on 8 key areas:

• Cash Flow (CF): Change in how well the company manages its cash

• Efficiency (EF): Change in how well they use their resources

• Liquidity (LIQ): Change in how easily they can pay short-term bills

• Profitability (Prof): Change in how much money they're making

• Solvency (Solv): Change in how well they can handle their debts

• Valuation (Value): Change in how fair their stock price is

• Composite (Comp): Combines all 6 measures into one number

• Factor (Fac): A measure of statistical change (can ignore)

We supply 650+ of the the largest positive and negative moving ticker ratios from the week.

• ZBAO demonstrated strong financial health with high scores across multiple metrics, particularly in CF (66) and Liquidity (12), outperforming peers in overall financial stability.

• MQ showed concerning financial metrics, with notably weak scores in Profitability (-25) and Solvency (-14), indicating deteriorating financial health compared to industry peers.

• Key Trends to Watch:

  • Positive: Companies improving in both Cash Flow and Profits (top right of the plot) often make great investments—they’re getting better at making and retaining money.

  • Negative: Companies struggling with both Cash Flow and Profits (bottom left) indicate potential issues.

Insider Flow Pressure

This model forecasts insider trading behavior, the idea is to predict when insiders will buy and sell stocks and to ‘front-run’ them. This model self-improves over time using machine learning. The tables focus on the top 20 stocks with the highest buying pressure and the top 20 stocks with the highest selling pressure.

Shows how a stock's predicted pressure compares to similar companies, ranked as a percentage

• NTRS shows the most promising flow prediction indicating a potential buying opportunity within the Financial Services sector.

• NKLA has the most concerning flow prediction suggesting significant selling pressure in the Industrials sector.

Looking at the change in a prediction from one week to another gives you a second-order output that is often a better leading indicator to stock price movements.

• AYI shows a promising change in flow prediction of 36.18, indicating a potential buying opportunity within the Industrials sector.

• AXP shows a concerning change in flow prediction of -80.06, suggesting significant selling pressure in the Financial Services sector.

Short Interest Analysis

We identify stocks whose short interest significantly deviates from comparable companies, refreshed every two weeks. The table focuses on the top and bottom 60 overshorted stocks.

• ELAB is heavily overshorted (+12.5% change) with 78.8% vs expected 27.2% - potential squeeze candidate.

• VSTE shows declining short interest (-23.6% change) at 3.3% vs expected 67.4% - this is bearish given that future shorts could catch on.

By the time the short-interest file becomes available, the market may have already reacted. Monitoring changes in short volume, however, could provide an earlier indication of shifts in short sentiment. Displays top and bottom 150 stocks by short volume ratio changes.

• IPHA shows rising short volume (+81.1% change) to 98.9% vs previous 17.8% and is a potentially concerning trend.

• CNL has declining short volume (-70.5% change) to 20.1% vs previous 90.5% and could signal covering with upside potential.

Wikipedia Search Pressure

This model uses an algorithm to discover early search interest on Wikipedia pages. This signal is stronger than most retail indicators as it suggests the beginning of a research process. We deliver the 100 tickers with the largest and smallest predicted changes in search behaviour.

• Interest in DOCS, based on Wikipedia searches, increased 32.04% this week, showing strong positive attention.

• Meanwhile, TAL experienced a significant decline of -27.57% in search interest, indicating reduced market attention.

Sentiment Modelling

Tracking news-based sentiment for individual companies allows us to aggregate sentiment to the sectorial level. We first deliver the 280 tickers with the largest and smallest predicted changes in sentiment. The aggregated data is then presented on a sectorial and thematic basis.

• News sentiment for JBL articles was most positive this week with a 59.009 change in sentiment score, suggesting favorable coverage.

• Meanwhile, ACRS faced the most negative news sentiment shift of -60.000, indicating challenging media coverage.

The chart above tracks the sentiment at a sectorial level which is helpful to assess weekly sectorial sentiment shifts. An increasing percentile indicates improving sentiment for a given sector/theme relative to others.

We also track numerous themes that we aggreagate in a handful of main themes, the detailed themes are available in the SDK. An increasing percentile indicates improving sentiment for a given sector/theme relative to others.

Breakout Prediction

Using machine learning models, we track the predicted change in price breakouts and slope. The implication of a change is that a stock has become more or less likely to go up in price. We deliver the 80 tickers with the largest and smallest predicted changes in breakout and slope.

• AFBI shows the strongest breakout prediction signal, suggesting potential upward movement.

• CABA displays the weakest breakout prediction signal, indicating possible downward pressure.

• UPS demonstrates the largest positive change in breakout prediction, suggesting improving momentum.

• MAT shows the largest negative change in breakout prediction, indicating deteriorating momentum.

• BBDC exhibits the steepest positive change in breakout slope, suggesting accelerating upward momentum.

• JJSF shows the steepest negative change in breakout slope, indicating accelerating downward pressure.

Congressional Trading Tracker

This tool tracks filings and records in both the House and Senate, based on the premise that members of Congress may be more informed than the average investor. The table shows trades performed in a given week.

Consumer Financial Complaints

We investigate recent changes in risk for financial firms based on complaints received in the latest week available. We deliver all the tickers that experienced a complaint this week with their change in risk. Please confirm the ticker-company name label, the mapping is still in beta and will improve over time.

The total risk rating is an ML-derived score measuring potential company liability based on complaint severity rather than volume.

• NYCB received 45 consumer complaints in the preceding month, with a concerning rise in financial risk of 6.21 points and a notably high culpability score of 88.55.

• Despite 40 consumer complaints, SLM shows significant improvement with financial risk dropping 15.07 points, while maintaining a customer grievance score of 14.34.

At Sov.ai Research we empower investors with rich data and analytics. Our offerings include insightful newsletters and an easy-to-use Python SDK. Enjoy this week’s newsletter! This newsletter content is for informational purposes only and should not be considered investment advice. The views and opinions expressed here are those of the author and do not constitute financial or investment recommendations. Always conduct your own research and consult with qualified financial advisors before making any investment decisions.

Government Spending

This data tracks recent government contract announcements. Given that the government accounts for over 40% of national expenditures, these contracts are an important source of revenue for many companies and are a reliable predictor of quarterly revenue. We deliver all the tickers with new government contracts for the associated week.

• Size indicates the size of the contract in USD millions. Obligation is the contractual amount the government ought to pay the company.

• Awards corresponds to the number of contracts awarded in the associated week. Previous indicates the number of contracts previously awarded to the same company.

Corporate lobbying tracker

Lobbyists have important roles to play in the American political system and are often responsible for the passage of bills and implementation of policies.

Anomaly corresponds to a change in lobbying behavior compared to historical lobbying patterns. It could indicate that it is a companies first lobbying contract, or a large contract compared to their historical average.

Earnings Surprise Prediction

Here we are interested in what companies are expected to experience an earnings surprise in the near future. The benefit of predicting earnings surprises is that you can act before the market reacts.

• The table only shows company that are reporting in the next two weeks from the date indicated at the top of the table.

• The earnings surprise predictions are bounded between -50% and +50% to filter out statistical outliers and maintain focus on meaningful variations.

• IMMX is predicted to potentially outperform expectations with a 49.7% likelihood of earnings surprise. The consensus EPS estimate is $-0.16 for the quarter ending 2024-09-30.

• LYV shows potential downside risk with a -50.5% earnings surprise probability. Analysts expect EPS of $1.60 for the quarter ending 2024-09-30.

Factor Model Coefficients

This analysis shows percentile rankings of factor model coefficients across stocks. Higher percentile coefficients indicate stronger factor sensitivity relative to peers. We deliver approximately 1000 tickers and their associated factor scores

Factor Model Error Analysis

Stocks with high R², negative AIC, and strong t-statistics provide reliable factor exposure for long-term systematic investing, while those with lower statistical significance and model fit offer alpha potential through market inefficiencies but demand sophisticated trading strategies.

Thank you for reading and subscribing. We aim to add a new screen every month and make it available to you on Mondays.

At Sov.ai Research we empower investors with rich data and analytics to build a thriving investment ecosystem. Our offerings include insightful newsletters and an easy-to-use Python SDK.

• Exclusive Newsletter: Experience a newsletter that delivers machine learning-driven predictions across a universe of 5,000 U.S. publicly traded stocks.

• Advanced Models & Data: Harness the power of over 20 primary source datasets and 30 state-of-the-art models from both academic and industry research.

• Affordable Pricing: Introductory rate at $48/month and founder subscription available for $73/month.

• Significant Savings: Founder subscription offers a 65% discount to the original sov.ai subscription ($285/month).

• Free Access Option: All newsletters are freely available to non-paying subscribers with a 2-week lag.

Founder Access: For $73/month you get access to the (1) newsletter, (2) and all datasets and Python analytics via the SDK at a 65% discount to the original sov.ai subscription of $3,420/year.