Que demande le peuple ?? 
Tu m’as convaincu… y a plus qu’à s’envoyer les 950 pages !! 
4 « J'aime »
Man Group étudie près de 40 conflits survenus depuis la Seconde Guerre mondiale afin d’analyser leurs effets habituels sur les marchés et les stratégies de protection des investisseurs. Les actions se redressent généralement, sauf si un conflit provoque un choc énergétique ou s’accompagne d’une crise de crédit. L’or, le pétrole, les taux d’intérêt et le dollar réagissent quant à eux différemment.
2 « J'aime »
Salut les geeks,
Bien que partisan de l’approche passive, je pense qu’il ne faut pas la transformer en dogme et ne surtout pas mettre son cerveau sur pause. C’est pourquoi j’ai beaucoup apprécié l’article “Passive Aggressive: The Risks of Passive Investing Dominance” (2025) de Chris Brightman (Research Affiliates, LLC) et Campbell R. Harvey (Duke University). ), disponible dans la langue de Shakespeare.
Les deux auteurs montrent que la progression fulgurante de l’investissement passif (via essentiellement les fonds indiciels pondérés par la capitalisation boursière) pourrait créer des distorsions de marché. Sa domination tendrait à éroder la diversification (les flux étant canalisés vers les mêmes entreprises), à détériorer le price discovery et à accentuer les bulles. Ils ne prônent pas un retour à la gestion active traditionnelle, mais plutôt des stratégies équipondérées (equal-weight), ou pondérées par les fondamentaux (ventes, cash-flow, book value, etc.)… Ce serait peut-être l’occasion d’avoir vos avis sur l’équipondération ?
Bonne journée,
Rémy
P.S. Article un peu hors sujet ; post publié aussi dans un autre forum financier où il n’ a pas suscité beaucoup d’intérêt
Salut Schtroumpf Remy,
Voila deux vidéos qui peuvent t’interesser, il y a des sources dans la description (que je n’ai pas lu)
Concernant la question la distortion du marché et le price discovery :
The Index Fund Bubble - YouTube
Concernant l’équipondération :
The Problem with Equal Weight Index Funds - YouTube
Tu me pardonnera de ne pas te faire un résumé car je risque de déformer 
1 « J'aime »
Salut Mmmh (bravo pour le pseudo)
Merci ! J’avais déjà vu les vidéos de Ben Felix, que je trouve très pertinent ! J’encourage les autres forumeurs et forumeuses à les consulter, en plus c’est bon pour notre anglais 
En défense de l’investissement passif, il y a ce papier intéressant de Vanguard : Setting the record straight: The truths about index fund investing (August 2025)
En bref, selon Vanguard:
- les fonds indiciels ne représentent pas une part excessive de la capitalisation de marché :
- Les fonds indiciels englobent des stratégies de plus en plus diversifiées :
- Les fonds indiciels ne détiennent pas une part excessive de grandes capitalisations :
- Les volumes d’échange en fonds indiciels sont une goutte dans l’océan :
Tout le reste est à l’avenant…
Auparavant, j’étais assez convaincu par ces arguments et me sentais même assez Bogleheads dans l’esprit. Maintenant, je trouve que c’est un peu comme si un chaîne de boucheries produisait toute une série d’études prouvant que la viande rouge n’est pas cancérogène… J’ai donc quelques réserves.
Bon week-end 
1 « J'aime »
Si vous vous posez la question pourquoi le marché japonais continue de monter un bon papier ici.
EN quelques mots (traduction)
« Le gouvernement japonais a donc tout intérêt à maintenir la vigueur du marché boursier national. Les rendements des actions japonaises semblent apporter une solution directe aux problèmes budgétaires du pays. Le secteur public détient des actions nationales représentant 41 % du PIB en 2025, réparties entre les fonds de pension publics (14 %), la Banque du Japon (13 %) et d’autres institutions publiques. Si les entreprises japonaises améliorent leur rentabilité et que la valorisation de leurs actions augmente, le patrimoine de l’État se renforce, parallèlement à ses recettes fiscales ».
« Lorsque l’État détient une part importante du marché et se finance à moindre coût, la hausse des cours boursiers n’est pas seulement avantageuse pour les actionnaires ; elle renforce également le bilan de l’État. En ce sens, Tokyo ne se contente pas de réformer le monde des entreprises japonais. Elle gère l’un des plus importants portefeuilles d’investissements nationaux à effet de levier au monde. »
A noter que I’ indiciel Japon 225 a toujours été inclus en continu dans le top 4 des assets sélectionnés par ma strat HAA (hybrid asset allocation) depuis mai 2025 avec un retour mensuel moyen de 4.3%, meilleur mois .
3 « J'aime »
Un papier super intéressant de Jack Vogel commenté par Wes Gray ( les 2 sont chez alpha architect )
"De nombreux investisseurs factoriels (factor investing) connaissent bien l’investissement « valeur des petites capitalisations (plus parlant en anglais small-cap value investing,”)», une stratégie d’allocation judicieuse pour les investisseurs à long terme capables de tolérer une forte volatilité.
Pourquoi observe-t-on un tel engouement pour les investisseurs axés sur la valeur des petites capitalisations ?
On a souvent dit à ces investisseurs que la prime de valeur est plus élevée, en moyenne, pour les petites capitalisations que pour les grandes. Malheureusement, cela s’avère faux pour un investisseur « long-only » . Notre[ collègue Jack Vogel a récemment publié un article intitulé « Investissement valeur long-only : la taille ?» qui démontre clairement ce point .](https://small%20value%20investing,-728r)
Le fait que les actions de valeur à grande et à petite capitalisation affichent des rendements similaires peut surprendre les investisseurs".⁹
1 « J'aime »
Un papier super intéressant sur un sujet relativement peu abordé et une bonne explication du base rate de Mauboussin.
Désolé del e copier en entier ici. Pas moyen de trouver un lien.
**Evidence-Based Forecasting Techniques for the Average Investor
« Investments of every kind are explicitly or t a bet on the future of asset prices. » -Brock et al.
| Investing is, at its core, an exercise in forecasting. Every DCF model, every portfolio optimization, every capital allocation decision ultimately rests on predictions about the future. What will different asset classes return? What will EBITDA growth be over the next five years? Will valuation multiples mean revert? To outperform, investors ultimately need to make better forecasts than are already embedded in prices. And yet there is a strange disconnect. Investors spend years studying accounting, finance, economics, and valuation, but remarkably few spend any time studying forecasting itself. We routinely make probabilistic judgments about uncertain futures without ever asking what the science says about how forecasts should be made—or why some people consistently forecast better than others. That science has advanced enormously over the past four decades, driven largely by the work of Philip Tetlock. Tetlock’s research has transformed forecasting from an art into something much closer to a science. If investing is fundamentally a forecasting exercise, then every serious investor should understand what the world’s best forecasters do differently. Over the past two decades, forecasting has become a much more rigorous discipline. In practice, nearly every major advance answers one of four questions: |
|---|
- What usually happens? — base rates
- What does new information imply? — prediction markets
- What have I overlooked? — AI and dialectical reasoning
- How can I reduce random judgment error? — aggregation
We’ll discuss each in turn.
Michael Mauboussin has been a champion of the use of base rate forecasting in investing. Two years ago, he applied these techniques to forecasting AI revenues. In 2024, OpenAI had revenue of $3.7B and was forecasting 2029 revenue of $145B. Mauboussin built a comparable historical distribution to see where OpenAI’s forecast landed.
Figure 1: Base Rates of 5-Year Sales Growth for Firms with $2-$5B in Sales, 1950-2024
| Source: Mauboussin By compiling a distribution curve that displays the most frequent five-year sales growth rates for 18,897 comparable firm period observations, he demonstrated that no proxy company has grown as fast as OpenAI’s projected rates (108% CAGR over five years). The lesson is broader than AI. Whenever investors face an uncertain forecast, the first question should be, “What happened in comparable situations before?” The mistake investors most commonly make is beginning with the inside view—this company, this management team, this industry—and only later asking whether the implied forecast is historically plausible. Evidence-based forecasting reverses that sequence. We think this type of base rate forecasting should become standard practice in finance. About a decade ago, we promoted a simple approach for using base rates to build DCF models. We argued that revenue growth forecasts should be accompanied by low confidence and humble confidence intervals, that margins typically compress over time, and that valuation multiples are a mean-reverting time series with confidence lowering as time intervals expand. Psychologists Daniel Kahneman and Dan Lovallo showed that people systematically underweight distributional information in favor of vivid, case-specific details. Their prescription was the “outside view:” Begin with historical base rates before incorporating facts unique to the situation. Mauboussin created a base rate book to help investors who need a set of “outside view” data to incorporate into their processes. Equally important, not all base rates are equally informative. Some variables, such as valuation multiples and profit margins, display substantial mean reversion. Others behave much more like random walks, where historical averages provide relatively little guidance. The forecaster’s first task is therefore to identify the appropriate reference class and determine whether history is likely to repeat itself at all. Choosing the wrong base rate can be worse than having no base rate. Base rates are the indispensable first step, but they are not the last. We now turn to the remaining three techniques. We believe prediction markets are one of the most useful innovations in forecasting since Tetlock’s original work. These markets fulfill Tetlock’s challenge two decades ago of adding quantification to a field formerly characterized by vague language and unchecked pundits. Prediction market–based forecasts consistently provide better accuracy than survey-based ones, reducing foresight error by an average of 5.5%. As such, Kalshi and Polymarket offer useful probability estimates in the short term. Investors should think of prediction markets the same way they think of Treasury yields or futures curves: not as infallible forecasts but as continuously updated market-implied probabilities. If your investment thesis depends on a macroeconomic variable or binary event that is actively traded on a prediction market, that market provides an objective benchmark against which to compare your own view. When your estimate differs substantially from the market’s, you should have a clear reason why. Additionally, prediction market time-series data reveals how heavily new information impacts public sentiment. Each question yields a graph of probabilities over the course of weeks and months. For macro variables such as unemployment rate, core CPI, and headline CPI, very short-term Kalshi estimates display an accuracy similar to the Bloomberg Consensus. Indeed, for headline CPI, Kalshi probabilities had either no statistically significant difference from, or “a significantly smaller mean absolute error [than,] the Bloomberg Consensus.” As such, if a short-term financial metric is impactful to their forecast, or if their forecast aligns closely with a question that exists on a prediction market, forecasters may want to benchmark against the wisdom-of-the-crowds probability. Additionally, prediction markets tend to incorporate new information expeditiously. A forecaster looking to accurately update their foresight can look to prediction markets to see how heavily recent news has impacted public outlook. AI is perhaps the most promising new forecasting tool, but not because it possesses superior judgment. Rather, its value lies in enforcing a disciplined forecasting process. Left to our own devices, we tend to jump to conclusions, overweight recent information, neglect base rates, and stop searching once we find confirming evidence. Properly prompted, AI can systematically counteract these biases. We surveyed the recent forecasting literature and found that the best prompts ask AI to mimic the process used by Tetlock’s superforecasters by beginning with base rates. Rather than asking directly for a prediction, we first ask the model to identify the relevant historical base rates, retrieve evidence from credible sources, and reassess their initial judgements by constructing arguments both for and against the proposition. The value of AI is not that it replaces judgment but that it imposes structure, helping the forecaster to think through prediction using a scientific framework. Perhaps the most useful application is often not generating an initial forecast but challenging one, asking the model to identify the strongest arguments against your conclusion. This reduces overconfidence while preserving independent judgment. One can also internally imagine reasons why one’s forecast might be wrong, then produce a second probability estimate that one then averages with the first estimation. This process is known as dialectical bootstrapping, and it tends to increase estimation accuracy by 4.1%. Overconfidence and idiosyncratic judgement error can be consistently, effectively reduced through probability aggregation. Indeed, though Tetlock was first to identify a number of foundational forecasting techniques, one of his most impactful insights is that teams of well-calibrated forecasters making independent estimates are much more accurate than any individual by themselves. Forecasters who cultivate a “superteam” of their own by averaging their estimates with a group of colleagues—or even just one independent individual—will, on average, raise the accuracy of their estimates by 7.1%. The future is riddled with randomness and uncertainty. A well-reasoned judgement might turn out wrong due to unforeseeable circumstances; a poor forecast might turn out to be right for unanticipated reasons. However, by systematically reducing cognitive biases, quantifying forecasts, aggregating opinions, and learning retroactively from mistakes, the average investor can improve the calibration of their predictions and boost the visibility of outcomes on the hazy horizon. Before making any important forecast, investors should ask four questions. What do the historical base rates imply? How is this currently priced in forecasting markets? What is the strongest argument against my conclusion? And whose independent forecast can I average with my own? None of these techniques eliminates uncertainty, but together they produce forecasts that are measurably better calibrated than intuition alone. Since every investment is ultimately a forecast, learning to forecast better may be one of the highest-return investments an investor can make. Acknowledgment: This piece was authored by our summer intern, Alexa Burton. Alexa is a rising junior at Brown University pursuing a double concentration in English and Behavioral Decision Sciences. In her free time, Alexa is a long-distance runner and a screenwriter for Brown’s short film organization. Next summer, she will be joining Goldman Sachs Asset & Wealth Management in their New York office. |
|---|
2 « J'aime »
Peut-être une source d inspiration pour l equipe de cayas pour le parcours pédagogique.
Tips, tricks, and activities for teaching economics using live data from FRED.
Ici
FRED c est des data financières en libre accès de la fed.
Tips, tricks, and activities for teaching economics using live data from FRED.
Visualizing the components of investment
This assignment provides instructions on building the graph below and includes suggested writing prompts for out-of-class assignments.
From the FRED Blog
The post “Negative investment?” uses data from the U.S. Bureau of Economic Analysis to compare gross with net investment.
Quiz yourself on investment
Interpret graphs to answer questions about investment. These questions reinforce economic analysis and data literacy skills. Share the FRED dashboard with your students.
Subscribe to this newsletter and a variety of research, educational, and historical content from the St. Louis Fed.
1 « J'aime »