Investieren und Märkte

Worum es geht

An asset price is a claim on uncertain future cash flows, and it equals the expectation of those cash flows discounted at some rate. Everything contested in the subject follows from the two unobservables in that sentence. Since Bachelier in 1900 it has been known that price changes look close to random; Samuelson showed in 1965 that this is what competition among informed traders would produce, and Fama in 1970 organised it into the efficient market hypothesis and its weak, semi-strong and strong forms. Markowitz in 1952 made diversification a covariance calculation and separated the risk that can be removed from the risk that cannot; Sharpe in 1964 turned that separation into a pricing statement, and the empirical failure of that statement produced the factor literature, its replication problems, and no successor everyone accepts. Against this, Shiller in 1981 showed that prices move far more than the subsequent path of dividends can justify, De Bondt and Thaler showed that extreme past performance reverses, and the limits-of-arbitrage literature explained why obvious mispricings can persist. The two readings of the same evidence — time-varying rational risk premia against systematic mispricing — are what the 2013 prize honoured together. Underneath the dispute sit facts neither side denies and both find inconvenient for the popular story: costs compound against the investor with certainty, the average actively managed dollar must underperform the average passive dollar after costs by arithmetic, most individual stocks underperform Treasury bills over their lifetimes, and the normal distribution understates extreme moves badly enough to have broken the risk models of 2008.

Was im Paket steckt

Kuratiertes Gerüst, das dein Agent vor der Recherche lädt, damit er auf geprüftem Boden startet statt bei null.

• Tier 1: 112
• Tier 2: 14
• Tier 3: 11
• Tier 4: 4

Die Fragen und Behauptungen unten sind wörtlich aus den Paketdateien zitiert. Die Pakete sind durchgehend englisch, weil sie für Agenten geschrieben sind.

Wo das Feld sich uneinig ist

Jede trägt echte Vertreter auf mehr als einer Seite, damit dein Agent nicht still einen Sieger kürt.

• Are asset prices rational?

  5 benannte Positionen · Genuinely unresolved, and the discipline said so institutionally by awarding the 2013 prize to Fama, Hansen and Shiller together, with a background document that states the interpretation of long-horizon predictability is open. Teach both readings at full strength, teach the joint hypothesis as the reason the dispute cannot be settled by pointing at data, and label any confident verdict — in either direction — as contested.

• Can the efficient market hypothesis be tested at all?

  4 benannte Positionen · The joint-hypothesis problem is not disputed by anyone; what people dispute is what it implies. The productive stance is to treat efficiency as a benchmark whose deviations are measured for a purpose, which is what Fama's own late writing says, rather than as a proposition to be affirmed or denied.

• Are the value and momentum premia compensation for risk or evidence of error?

  4 benannte Positionen · Unresolved and probably not resolvable by the same data, since the two accounts predict the same average returns. The distinguishing predictions concern behaviour after publication, behaviour in different institutional settings, and correlation with plausible bad states, and the evidence there is mixed.

• How many published factors are real?

  4 benannte Positionen · Live and methodological. The raw data are shared; the disagreement is about statistical protocol, which is why it has not converged. The defensible working position is that a small number of factors — market, value, momentum, profitability, low risk — have survived the most scrutiny, that the long tail is unreliable, and that any specific factor claim should be checked against the current replication literature rather than against its original paper.

• Can active management add value, and for whom?

  6 benannte Positionen · The arithmetic is not in dispute and the boundary conditions on it are real but narrow. The empirical record on net returns is one of the more consistent findings in finance. What remains genuinely open is whether specific investors can identify skill in advance, and the honest answer is that no published method does so reliably.

Mythen, die das Paket korrigiert

Verbreitete Behauptungen mit der Evidenz, die sie klärt oder begrenzt.

• “A fund, manager or strategy with a strong recent record is likely to continue outperforming, so past returns are the sensible basis for choosing one.”

  debunked-as-stated

  Persistence in net returns is weak. What persistence exists at short horizons is largely attributable to momentum exposure and to expenses rather than to skill; the most reliable persistence is at the bottom of the distribution, where high costs keep poor funds poor. Bootstrap analysis of the whole cross-section finds that few funds produce benchmark-adjusted returns sufficient to cover their costs, and false discovery methods find the proportion of genuinely skilled funds small and declining. Survivorship correction matters enormously: closed and merged funds leave the database, and survivorship alone can manufacture apparent persistence out of pure noise. There is a kernel of truth to preserve — Hendricks, Patel and Zeckhauser did find short-horizon persistence, and Carhart then showed what most of it was — and there is an interpretation that unsettles the tidy conclusion: Berk and Green show that if skill is real but scarce, and money chases it until scale exhausts the edge, then zero net alpha to investors is exactly what a rational market containing skilled managers produces. The regulatory disclaimer that past performance does not predict future results is not legal throat-clearing; it is one of the better-evidenced statements in the field, and one cost is entirely predictable while returns are not.

• “The efficient market hypothesis says prices are always correct and investors are rational, so 2008 and every other crash refuted it.”

  debunked-as-stated

  Fama's claim is that prices reflect available information — a statement about what is knowable, not about what is correct. Prices can be wrong in hindsight and still be efficient, because efficiency means no one could have known better at the time. The hypothesis comes in three nested forms by information set: weak (past prices), semi-strong (all public information) and strong (all information including private). The strong form is known to fail and Fama treated it as a benchmark rather than a description; insider trading is profitable, which is why it is illegal. Nor does the hypothesis require rational investors: it requires that errors do not leave systematically exploitable patterns, which is a much weaker condition. Then comes the part that decides most arguments: the joint-hypothesis problem. To call a return abnormal one needs a model of normal returns, so every test of efficiency is a joint test of efficiency and of that model, and a rejection never says which failed. Fama stated this in 1970, concluded in 1991 that market efficiency per se is not testable, and repeated it in his Nobel lecture. Roll made the specific version: the true market portfolio is unobservable, so tests of the capital asset pricing model test the proxy. Two further corrections. The hypothesis does not claim markets are perfectly efficient — Grossman and Stiglitz proved that state is impossible, since no one would pay for information no price rewarded. And Fischer Black, from inside the tradition, proposed that an efficient market is one where price is within a factor of two of value most of the time, which is a long way from "always right". Crashes are therefore not automatic refutations, and the joint-hypothesis knife cuts both ways: it protects efficiency from refutation and equally protects anomalies from being explained away.

• “With enough attention to indicators, valuations or news, an investor can reliably move in and out of the market ahead of major moves.”

  debunked

  Three independent lines of evidence point the same way. First, forecasting records: Cowles audited thousands of professional recommendations in 1933 and found no demonstrable skill, and the design has been repeated since with the same result. Second, the arithmetic of the required accuracy: a timing strategy must be right often enough to overcome transaction costs, taxes on realised gains, and the expected return forgone while out of the market, and the accuracy threshold this implies is high. Third, what investors actually do: dollar-weighted returns, which account for when money entered and left, are materially below buy-and-hold returns across US and international markets, and individual investors' net returns fall monotonically with turnover. The familiar demonstration that missing the best few days destroys most of the long-run return is a real illustration of how concentrated returns are, but it is a weak argument by itself and should be taught with its symmetry: the best days cluster in the same turbulent periods as the worst days, and the calculation for missing the worst days is rarely shown alongside. The correct statement is not that markets are unforecastable in every sense — valuation ratios do forecast long-horizon returns, which is one of the few things Fama and Shiller agree on — but that a probability shift over a decade is a different object from a signal to move money this quarter, and treating the first as the second is the error.

• “A properly diversified portfolio has removed its risk, so a diversified investor should not lose money in a downturn.”

  debunked-as-stated

  Diversification removes idiosyncratic risk — the part specific to one company, sector or country — and it does not remove systematic risk, the part shared with everything else. The variance curve flattens toward market risk, not toward zero, and a fully diversified equity portfolio still carries the whole of the market's risk. This is not a defect of the technique but the foundation of asset pricing: only the undiversifiable part earns compensation, precisely because the diversifiable part can be removed for free. Two further corrections make the point practical. The correlations that diversification depends on are estimated from history and are not stable: correlation between equity markets and between portfolios rises in extreme downward moves and does not rise symmetrically in extreme upward moves, a distributional property established with extreme value methods rather than an anecdote about one crisis. So diversification delivers least at the moment it is most wanted. And the number of holdings required is not a constant: firm-level volatility has varied substantially over time, so any fixed rule about how many positions constitute a diversified portfolio is dated the moment it is quoted. What diversification does deliver is worth stating positively, because the correction can be over-read: given how skewed long-run individual stock returns are, with most stocks underperforming Treasury bills and the aggregate gain traceable to a small minority, breadth is the only reliable way to hold the few that mattered.

• “In 1637 the Dutch traded houses for tulip bulbs, and the collapse of the bulb market ruined fortunes across society and devastated the Dutch economy.”

  debunked

  Anne Goldgar's archival work in Dutch notarial, court and estate records found no support for the collapse the story requires. The trade was concentrated among a limited network of merchants, skilled artisans and connoisseurs in a small number of towns, not a national frenzy across classes. The extraordinary prices attach to a small number of rare varieties, and the volume of trade at those prices was small. The market operated largely in forward contracts due at bulb lifting; when bidding collapsed in February 1637 most contracts had never been paid, enforcement was suspended, and disputes were later settled at a small fraction of contract value, so the paper losses were mostly never realised. Goldgar found no evidence of mass bankruptcies, no documented suicides attributable to the episode, and no economy-wide damage; the Dutch Golden Age continued. What the crisis actually was, and this is more interesting than the myth, was a crisis of trust and honour in a society where credit was personal and reputation was collateral. Two other revisions exist and should be kept distinct from Goldgar's, because collapsing them is how a correction becomes the next myth: Peter Garber argues from prices that rare-bulb valuations were broadly consistent with the fundamentals of a slowly propagated luxury good, and Earl Thompson argues that the late price spike reflects a change in the legal character of the contracts toward option-like obligations rather than a change in belief. Goldgar's claim is about consequences and is archival; Garber's claim is about rationality and is contested; they are not the same claim. The general lesson is the one worth carrying: a historical anecdote repeated in every book of a genre is evidence about the genre, not about the past, and the question to ask of any such story is who checked.

Verwandte Themen

• Wirtschaftspsychologie
• Boom-und-Bust-Zyklen
• Geldgeschichte
• Spieltheorie