Portfolio Lab

Glossary: terms starting with A

52 terms from Abnormal return to Axe, briefly explained and with a source where possible. All terms from A to Z

Abnormal return (CAR, Cumulative abnormal return, Event study)

The actual return minus the normal, expected return, usually from a market model (α + β × market return). Summed over several days, this gives the cumulative abnormal return (CAR), the standard measure in event studies.

Example: With a beta of 1.2, a stock return of 1.5% and a market return of 0.5%, the abnormal return is 1.5 − 1.2 × 0.5 = 0.90%.

Source: Fama et al. (1969)

More: Insider trading: what is prohibited, what is allowed and what insider purchases reveal

Absolute priority rule (Absolute Vorrangregel, APR, Priority rule)

Principle that in insolvency and restructuring, a junior creditor or owner receives nothing as long as a senior one has not been paid in full. In US law, it is part of confirming a plan over a dissenting class; in practice there are deviations.

Source: US Bankruptcy Code (1978); StaRUG (2020); Eberhart et al. (1990)

More: Distressed debt: strategies, bankruptcy law, recovery rates and returns when investing in crises

Absolute risk aversion (Arrow-Pratt measure)

Pratt's measure: A(w) = −u″(w) / u′(w). Indicates how strongly someone shies away from a fixed amount of money at risk. With logarithmic utility it falls with wealth, with quadratic utility it rises, which contradicts real behavior.

Source: Pratt (1964)

Absorption ratio (Kritzman absorption ratio)

The share of total variance explained by the few largest principal components. When it rises, markets move more closely together. This points to greater vulnerability to broad declines.

Source: Kritzman et al. (2011)

Methodology: Asset analysis: methodology from correlation to cointegration

Accruals anomaly (Accrual anomaly, Sloan accrual anomaly)

Companies whose earnings consist largely of accruals rather than cash flows have subsequently delivered lower returns on average. Such earnings are less persistent, and prices took this into account only with a delay.

Source: Sloan (1996)

More: Fundamental analysis: valuing a company like an investor

Accrued interest (Accrued coupon, Stückzinsen)

Interest that has built up on a bond since its last coupon date. The buyer pays it to the seller on top of the price, because the buyer receives the full coupon on the next date.

Example: 3% coupon, 226 of 365 days since the last coupon date (act/act): 3 × 226 / 365 = $1.86 per $100 face value.

Source: Tuckman and Serrat (2012)

More: Bond math: calculating price, yield, duration, convexity and spreads

Accumulating

A fund that reinvests its income instead of paying it out.

More: ETF basics: costs, replication and distributions

Active ETF (Actively managed ETF, Active exchange-traded fund)

Exchange-traded fund whose manager selects and weights securities at their own discretion instead of following an index. Usually cheaper than traditional active funds but more expensive than index ETFs; it too has to beat its benchmark after costs first.

Source: Easley et al. (2021)

More: Active ETFs: how they work and what they deliver

Active fund (Actively managed fund)

A fund whose manager selects and weights securities in order to beat a benchmark. Over long periods, most active funds lag behind their benchmark index after costs.

Source: French (2008); Fama and French (2010)

More: Active or passive investing: ETFs versus fund managers

Active return (Relative return)

The difference between the return of a portfolio and that of its benchmark. Its average divided by its volatility (tracking error) gives the information ratio.

More: Tracking error and information ratio: deviation from the index

Active risk budget (Tracking error budget, Risk budget for TAA)

How far a portfolio may deviate from its benchmark, measured as tracking error. It is divided between tactical allocation and selection; because variances add up, the budgets grow with the square root.

Example: If the return difference between equities and bonds has a volatility of 16.28%, a budget of 1% tracking error allows the equity weight to deviate by at most 6.1 percentage points.

Source: Grinold and Kahn (2000)

More: Strategic and tactical asset allocation and the total portfolio approach

Active share

Half the sum of the absolute weight differences between a fund and its index. 0% means a copy of the index, 100% no overlap at all. A high value shows active selection, but not yet skill.

More: Active or passive investing: ETFs versus fund managers

Activist investor (Activist, Activist shareholder, Activist hedge fund, Activist investing, Shareholder activism, Proxy fight)

Investor, often a hedge fund, that buys a minority stake and pushes for changes, publicly or behind closed doors: higher payouts, a new strategy, the sale of divisions or a change of management. If necessary, this happens via a proxy fight, a contested vote at the annual general meeting.

Example: According to Brav et al. (2008), the excess return around the announcement of a campaign in the US from 2001 to 2006 was about 7%.

Source: Brav et al. (2008)

More: Legendary investors: the investment styles of the most famous investors, from Graham to Simons

Ad hoc disclosure (Ad hoc announcement, Inside information disclosure, Ad-hoc-Mitteilung)

A mandatory announcement by an issuer under Art. 17 MAR: it must disclose inside information that directly concerns it without delay, and may delay disclosure only under strict conditions. Since June 5, 2026, for protracted processes such as takeovers, only the final event has to be disclosed.

Source: MAR (2014); Gleiss Lutz (2026)

More: Insider trading: what is prohibited, what is allowed and what insider purchases reveal

Adaptive Allocation

Tactical rule that each month picks the assets with the strongest momentum and then weights them by minimum variance or inverse volatility.

More: Tactical allocation: momentum, dual momentum and trend following

Adverse selection (Information asymmetry, Glosten-Milgrom model)

A market maker's risk of trading mainly with better-informed traders: it buys before prices fall and sells before they rise. The bid-ask spread compensates for these losses.

Example: Value 99 or 101 with equal probability, 20% informed orders: ask 100.20, bid 99.80.

Source: Glosten and Milgrom (1985)

More: Market making and bond trading: spread, inventory, axes, RFQ, cross trades and matrix pricing

Aframax (Aframax tanker, AFRA)

Medium-sized tanker with a deadweight of about 80,000 to 120,000 t, smaller than Suezmax and VLCC. Portfolio Lab classifies tankers from 230 to under 265 m in length this way.

Source: EIA (2014); Notteboom et al. (2026)

Methodology: Ships at chokepoints (beta): methodology

Agent-based model (Agent-based models, ABM)

Simulation of many simple traders who act according to rules and switch between strategies, such as following trends or betting on fair value. Produces fat tails and volatility clustering from their interaction, even when the news has no such patterns.

Source: Lux and Marchesi (1999)

AIS (Automatic Identification System, AIS transponder, Vessel tracking, Ship tracking, AIS data)

Automatic Identification System: radio system for safety at sea. Ships automatically transmit their identity, position, course and speed over VHF, along with name, type, dimensions, draft and destination. Under SOLAS, it is mandatory for ships of 300 GT and above on international voyages; the range at sea is typically 20 to 30 nautical miles.

Source: IMO (2015); ITU (2026)

More: Ship tracking data and satellite images: what they reveal about commodities · Methodology: Ships at chokepoints (beta): methodology

All Weather portfolio

A portfolio designed to be ready for every combination of growth and inflation: equities, long-term bonds, inflation-linked bonds, gold and commodities, weighted by risk rather than by capital.

More: All Weather portfolio: ready for every economic climate?

Allocation effect (Asset allocation effect)

The part of the active return that comes from differing segment weights. Under Brinson-Fachler it is (wᵢ − Wᵢ) × (bᵢ − B): an overweight pays off if the segment beats the overall benchmark.

Example: A 10 percentage point overweight in a segment returning 8% against a benchmark return of 6.8%: 0.10 × 1.2 = 0.12 percentage points.

Source: Brinson and Fachler (1985)

More: Performance attribution: Brinson decomposition, factors and linking over time

Alpha (Jensen's alpha, Jensen alpha)

The part of a fund's return that cannot be explained by its market risk (beta) or other factors. Positive means: better than the risk taken would have led you to expect.

More: Beta and alpha: how closely a fund moves with the market

Alpha decay (Signal decay, Post-publication decay)

The decline in a signal's return once it becomes known and more investors use it. Published return predictors delivered on average 58% less after publication than in the original sample.

Source: McLean and Pontiff (2016)

More: Alternative data: job ads, Reddit, X, satellites and card payments

Alternative data (Alt data, Alternative datasets)

Information about companies and markets beyond prices, financial statements and official statistics, e.g. job ads, social media posts, satellite images, card payments or location data. It is often available earlier, but has to be processed first and is frequently expensive.

Source: Zhu (2019)

More: Alternative data: job ads, Reddit, X, satellites and card payments

American option (European option, Early exercise)

Option that may be exercised at any time up to expiry, unlike a European option, which can only be exercised on the expiry date. For calls without dividends, early exercise never pays, but for deep in-the-money puts it can; the right to do so costs a premium.

Example: Put with price and strike 100, one year, 5% interest rate and 20% volatility: European 5.57, American 6.09.

Source: Merton (1973)

More: Option pricing: where fair option prices come from · Methodology: Option pricer: how Portfolio Lab values options

Amihud measure (ILLIQ, Amihud illiquidity, Amihud ratio)

An illiquidity measure based on daily data: the average ratio of the absolute daily return to daily trading value. The higher the value, the more one dollar of turnover moves the price; it can be calculated for any stock with prices and turnover.

Example: Returns of +1.2%, −0.8% and +2.0% on turnover of $2.0 million, $1.6 million and $2.5 million: (0.6 + 0.5 + 0.8) / 3 = 0.63% per million dollars of turnover.

Source: Amihud (2002)

More: Liquidity risk: measuring market liquidity, liquidity premiums and the tools of open-ended funds

Anchoring effect (Anchoring)

A random or irrelevant number, such as your own purchase price or a previous high, influences judgments. The purchase price says nothing about whether an investment is a good one today.

Annual increase (savings plan) (Contribution escalation, Dynamic savings plan, Step-up savings plan)

Increasing the contribution every year by a fixed rate, for example to keep pace with your salary. With an annual increase d, the monthly contribution in year k (counting from 0) is R × (1 + d)^k; over long periods, this raises the total amount paid in considerably.

Example: $150 a month with a 3% annual increase: in the eleventh year $150 × 1.03¹⁰ = $201.59 a month.

More: Compound interest: how returns on returns make wealth grow · Methodology: Methodology of the free calculators: FIRE, savings plan, withdrawal and leveraged ETF

Annualization

Scaling metrics up to one year: returns with compounding, volatilities with the square root of time (monthly values × √12, daily values × √252).

More: Volatility: how much a portfolio fluctuates

Anomaly

Return pattern that a standard model such as the CAPM does not explain, for example momentum or the low-volatility effect. Many anomalies weaken after they are published.

Source: McLean and Pontiff (2016)

More: The efficient market hypothesis: why it is so hard to beat the market

Anti-dilution levy (ADL, Dilution levy)

A fee charged only to the transacting investors on days with large inflows or outflows, which is paid into the fund. It is meant to cover the estimated trading costs so that the remaining investors are not diluted; unlike with swing pricing, the unit price stays the same for everyone else.

Source: IOSCO (2023)

More: Liquidity risk: measuring market liquidity, liquidity premiums and the tools of open-ended funds

Anti-dilution protection (Anti-dilution, Full ratchet, Weighted average anti-dilution)

Contract clause that protects earlier investors in a round at a lower price (down round): their conversion price is lowered, either all the way to the new price (full ratchet) or by a weighted average. The additional dilution is borne mainly by founders and employees.

More: Venture capital: funding rounds, valuation, the power law and what research knows about returns

Arbitrage pricing theory (APT)

Theory according to which expected returns depend on several common risk factors. It leaves open which factors these are and is the foundation of all multi-factor models.

Source: Ross (1976)

More: CAPM: the capital asset pricing model and the security market line

ARCH model (ARCH)

Autoregressive conditional heteroskedasticity: today's variance depends on the squared surprises of the previous days. The model captures volatility clustering; Robert Engle received the Nobel Prize for it in 2003.

Source: Engle (1982)

ARIMA model (ARIMA, ARMA, Box-Jenkins method)

A time series model following Box and Jenkins that combines autoregression (AR), differencing (I) and a moving average of past shocks (MA). For returns the estimated weights are usually small, and the forecast quickly reverts to the mean.

Example: AR(1) with φ = 0.10: a monthly return of 3% with a mean of 0.6% raises the forecast for the following month only to 0.84%.

Source: Box and Jenkins (1970)

More: Return forecasting models: ARIMA, GARCH, LSTM and machine learning

Arithmetic mean (Average return, Arithmetic average return)

The simple average of period returns. It overstates actual wealth growth as soon as returns fluctuate; on the other hand, it is the right measure for the expected return of a single period.

Example: +50% and −50% average 0%, but wealth falls to 75%.

More: Volatility drag: why volatility costs you return

Article 6 fund

A fund that makes no sustainability claim within the meaning of the SFDR. It only discloses whether and how sustainability risks are factored into its investment decisions.

Source: SFDR (2019)

More: ESG ETFs: index rules, ratings, returns and the limits of sustainable funds

Article 8 fund

A fund that promotes environmental or social characteristics, colloquially “light green”. The SFDR does not prescribe how strictly it has to select. Anything from a few exclusions to strict filters is possible.

Source: SFDR (2019)

More: ESG ETFs: index rules, ratings, returns and the limits of sustainable funds

Article 9 fund

A fund that has sustainable investment as its objective, colloquially “dark green”. If it has a climate objective, it must refer to an EU climate benchmark or explain how it will achieve that objective.

Source: SFDR (2019)

More: ESG ETFs: index rules, ratings, returns and the limits of sustainable funds

Asian option (Average option, Average price option)

Option on the average price over several fixings instead of the closing price. It is cheaper than the standard option and harder to manipulate; common for commodities and currencies. There is no exact formula for the arithmetic average.

Source: Kemna and Vorst (1990); Levy (1992)

More: Option pricing: where fair option prices come from · Methodology: Option pricer: how Portfolio Lab values options

Asset class

Group of investments with similar risk and return, such as equities, government bonds, corporate bonds, commodities, real estate or the money market.

Asset location

Which asset is held in which account when accounts are taxed differently, e.g. a tax-advantaged retirement account and a regular brokerage account. The same allocation, distributed more cleverly, can save tax.

Source: Dammon et al. (2004)

Asset swap spread (ASW, ASW spread, Asset swap)

The spread over a floating reference rate that a fixed-rate bond earns when its interest rate risk is exchanged via an interest rate swap. It measures the bond's credit risk and, alongside the Z-spread, serves as a point of comparison for the CDS premium.

More: Basis trades: positive and negative basis in bonds, CDS and futures

Asset-liability management (ALM)

Planning that sets assets against future payments: present value of the liabilities, funding ratio and required return.

More: Asset-liability management: aligning your wealth with future spending

Asset-referenced token (ART, Asset-referenced tokens)

A crypto-asset under MiCA that aims to maintain a stable value by referencing another value or a basket, e.g. gold, several currencies or crypto-assets (ART). The issuer needs authorization, a white paper and a reserve of assets; issues of up to €5 million on average over twelve months are exempt from authorization.

Source: BaFin, ART and EMT guidance notice (2025)

More: Digital assets: tokenization, stablecoins, crypto securities and the digital euro

Asymmetric correlation

Equities are more strongly correlated in falling markets than in rising ones. Diversifying across equity markets therefore offers less protection precisely in a crash than the average correlation promises.

Source: Longin and Solnik (2001); Ang and Chen (2002)

More: Correlation and diversification: why mixing assets reduces risk

Atomic settlement (T+0, Real-time delivery versus payment, Atomic DvP)

Settlement in which delivery and payment take place in a single, indivisible step: either both sides change hands or neither does. On a shared register this eliminates the settlement risk between trade and settlement; in return, cash and securities have to be in place beforehand.

Example: Share purchase of $10 million at 1.5% daily volatility: with T+2, the replacement risk at the 99% level is about $10 million × 2.326 × 1.5% × √2 = $493,419; with atomic settlement it is $0.

More: Digital assets: tokenization, stablecoins, crypto securities and the digital euro

Augmented Dickey-Fuller test (ADF test, ADF, Augmented Dickey-Fuller)

An extended Dickey-Fuller test for a unit root: the regression also includes lagged differences so that autocorrelation does not distort the test. A strongly negative test statistic points to a stationary series, such as a spread that reverts to its mean.

Source: Said and Dickey (1984); Dickey and Fuller (1979)

More: Pairs trading and statistical arbitrage: idea, methods, research · Methodology: Long/short and statistical arbitrage: methodology

Authorized participant (AP, Authorized participant)

An approved market participant, usually a large bank, that may create or redeem units directly with the ETF provider. It uses price deviations for arbitrage.

More: ETF basics: costs, replication and distributions

Autocorrelation

Correlation of a time series with a lagged copy of itself: is a good month more likely to be followed by another good one? Equity returns show hardly any autocorrelation; smoothed valuations (property funds) show a lot.

Availability payment (Availability-based PPP, Concession, Demand risk, P3, PPP, Public-private partnership)

Payment by the public sector to the operator of an infrastructure asset for keeping it usable, regardless of how many people use it; typical of public-private partnerships for roads, schools or hospitals. The counterpart is a concession with demand risk, where tolls or user charges determine revenue. Social infrastructure projects with such payments had the lowest default rate after ten years in Moody's data, at 1.1%.

Source: Moody's (2020)

More: Infrastructure investing: equity and debt, revenue models, valuation and research

Axe (Bid axe, Offer axe, Axed)

A dealer's specific interest in buying (bid axe) or selling (ask or offer axe) a security, usually because the dealer wants to reduce or cover a position. Under the ICMA standard the quantity has to be stated, while the price is optional.

Source: ICMA and FIX Trading Community (2025)

More: Market making and bond trading: spread, inventory, axes, RFQ, cross trades and matrix pricing

Brief, simplified explanations; not investment or tax advice.