Hinweis: Ich veröffentliche die Liste interessanter Studien hier mit einer Verzögerung. Die aktuelle Aufstellung erhalten Sie bei Anmeldung für meine Rundmail (kostenlos und werbefrei).
Social Media Algorithms as Information Intermediaries
While algorithmic curation could have the potential to improve investor outcomes by surfacing value-relevant content, our evidence suggests that, in practice, it directs attention toward immediately salient or sentiment-driven information, with adverse consequences for retail investors and overall market quality. […] Financial social media platforms are not passive conduits of user-generated content but active intermediaries that allocate investor attention.
Fazit: Beiträge auf Social Media, die vom Algorithmus hervorgehoben werden, sollte man lieber ignorieren.
AI meets DIY: The Impact of Human Intervention on AI-Assisted Investing
Across and within investor-AI conversations, greater investor input is associated with lower future Sharpe ratios. […] Investor input steers AI toward portfolios with higher past returns, higher past risk, and less diversification, consistent with return chasing and under diversification. […] Comparing portfolios investors choose to purchase with those they ignore, we find selected portfolios exhibit lower Sharpe ratios due to higher risk without improved returns. […] Sophistication does not mitigate the broader negative relation between investor input and performance.
Fazit: Privatanleger pfuschen gern dazwischen. Das wirkt sich negativ auf die risikoadjustierte Rendite aus.
How the SPIVA U.S. Scorecard Understates the Performance of Actively Managed Mutual Funds
We evaluate the actual returns of exiting funds instead of treating them as underperformers irrespective of their performance; we weight our results by fund assets instead of equal weighting; and we compare active fund performance against the actual performance of passive funds instead of against the performance of hypothetical benchmarks. After those changes, we find that active fund underperformance significantly decreases.
Fazit: Aktive Fonds performen nicht ganz so schlecht wie oft behauptet.
By reducing perceived difficulty and increasing confidence, LLMs increased stock market participation. […] LLM users adhere more strongly to initial choices even when presented with an opportunity to revise. […] LLMs amplify motivated reasoning. Taken together, these results document a fundamental tension in the adoption of generative AI. […] LLMs act as a behavioral accelerant: they significantly lower the barrier to entry for non-investors, yet simultaneously facilitate the entrenchment of motivated beliefs through confirmation seeking.
Fazit: Anleger müssen lernen, angemessen mit KI zu interagieren.
The Labubu market exhibits a complete speculative cycle in less than two years. […] Worldwide attention rose 41 times above its pre-catalyst baseline, peaked at the index ceiling, and declined persistently once the issuer expanded supply. […] Engineered scarcity, viral attention, and globally connected secondary trading combine to produce quasi-financial pricing dynamics in consumer products.
Fazit: Interessante Analyse des Labubu-Hypes und seines absehbaren Endes
Do Finfluencers Need Special Regulation? Market Impact of Finfluencers Under Liquidity Constraints
Finfluencers in frontier markets are not merely noise traders. They perform a structural role in market infrastructure. […] Positive finfluencer sentiment generates statistically and economically significant abnormal returns, whereas negative sentiment, dampened by short-sale constraints, fails to correct prices. Trading volume increases across all sentiment categories, up to 81.3% for positive posts and exceeding 290% for individual micro-cap stocks, confirming finfluencers‘ role as de facto liquidity providers.
Fazit: In Frontier Markets können Finfluencer die Kurse erheblich beeinflussen.
Generative AI and Investment Research: Evidence from Analyst Reports
AI-assisted reports are associated with improved forecast accuracy. […] Investors react more strongly to AI-assisted forecasts, indicating that markets perceive their underlying reports as incrementally informative. […] Efficiency gains from AI are more pronounced for less experienced analysts and for analysts facing higher workload demands. […] AI functions both as a productivity-enhancing tool and as a partial equalizer of skill differences across analysts.
Fazit: Eine Studie von vielen, die zeigen, dass KI aus dem Research nicht mehr wegzudenken ist.
Passive Flows and the Limits to Arbitrage
We use passive flow-induced trading to measure mechanical stock-level demand shocks created when index funds scale their holdings in response to investor flows. […] Passive inflows raise returns in both legs, but they raise short-leg returns by more, compressing long-short spreads exactly when arbitrageurs are betting against those stocks. […] Passive investing need not change aggregate active alpha mechanically to affect active management. It can instead change the risks, timing, and realized returns of the trades active managers use to exploit mispricing.
Fazit: Passive Kapitalströme können Arbitrage erschweren.
Hedge Fund Awards: Do Investors and Managers Care, and Should They?
Award-winning funds, along with non-winning nominees, attract substantial inflows following recognition, yet these flows are not justified by subsequent outperformance. Our evidence suggests that HFAs function primarily as marketing tools, directing investor attention and conferring external validation. […] Award-eligible funds – particularly those positioned as top contenders by midyear – increase return smoothing in the second half of the year.
Fazit: Hedgefonds-Investoren werden durch Awards an der Nase herumgeführt.
Fresh Data, Stale Positions: The Staleness Trap in Market Anomalies
This paper studies the within-portfolio position lifecycle of cross-sectional return predictors. Standard empirical asset pricing methods implicitly assume that sorting on the current level of a signal is a good proxy for sorting on the current return opportunity. This assumption is less reliable in market anomalies because signal persistence outlasts return persistence. Even when the underlying trading data are current, standard portfolio rules can retain aged positions whose economically relevant premiums have largely dissipated.
Fazit: Frische Signale funktionieren bei Renditeanomalien am besten.