40 terms
Plain-language definitions of the terms Acubic uses when it builds, tests and rebalances a portfolio, each linked to the article that covers it in depth.
Last reviewed
A
The rule that turns a chosen set of assets into target weights. Acubic offers optimizers (Hierarchical Risk Parity, mean-variance and CVaR) and simple rules such as equal weight, rank weighting and market-cap weighting, each inside the position limits of your risk level.
B
A simulation of how a rule-based portfolio method would have behaved on past data. It is a hypothetical result, not a record of real trading and not a forecast, and it ignores some real-world costs such as taxes and market impact.
One hundredth of one percent, so 25 basis points is 0.25%. Acubic backtests charge transaction costs as a flat number of basis points on the value of every buy and every sell.
C
The constant yearly rate that would take a starting value to an ending value over a period, with growth compounding each year. It smooths out the ups and downs, so it says nothing about the path taken or the risk borne on the way.
A ceiling on how much of a portfolio one asset may hold. Unless you set your own maximum, Acubic caps each asset at 10% for a conservative risk level, 15% for moderate and 25% for aggressive, widened for small portfolios so the weights can still sum to 100%.
A number from -1 to +1 describing how closely two assets' returns move together. Low correlation is what makes diversification work, and correlations tend to rise in sharp sell-offs, exactly when diversification is needed most.
The average loss in the worst slice of outcomes, for example the worst 5% of days. A CVaR optimizer, one of the allocation methods in Acubic, minimizes that tail loss instead of ordinary variance.
D
Spreading capital across assets whose returns do not move in lockstep, so that no single position decides the outcome. Holding many assets is not enough on its own if they are highly correlated.
A drawdown is a fall from a previous peak to a later low, measured in percent. The maximum drawdown is the largest such fall in the period tested, and it is a common way to describe the worst stretch an investor would have lived through.
E
An allocation that gives every asset the same share, so ten holdings are 10% each. It needs no forecasts, and it drifts as soon as prices move until the next rebalance.
A fund that holds a basket of assets and trades on an exchange like a share. Which ETFs you can buy depends on your broker and your region.
A security listed on an exchange that tracks an asset or index, a wider family than ETFs that also includes exchange-traded commodities and crypto products. When a broker does not list a particular asset directly, an ETP tracking it can be the way to hold exposure to it.
F
A measurable characteristic used to compare assets, such as momentum, Sharpe ratio, size or volatility. Acubic scores assets on factors at the selection stage and can blend several into one ranking.
An Acubic automation mode in which scheduled rebalances are submitted to the broker without a per-order approval, behind safety guards such as one order in flight at a time. On eToro it is available only on a sub-account; on Trading 212 it is available on the main account.
H
An allocation method that clusters assets by how their returns move together and spreads risk across the clusters. It avoids inverting a covariance matrix, which makes it less sensitive to noisy estimates than classical mean-variance optimization.
L
A minimum amount of trading activity an asset must have to be eligible. In Acubic an asset is ranked only if its median daily dollar volume over its last 60 trading days is at least $1 million, so portfolios are not built from assets that are hard to trade. It applies to strategies created or regenerated since October 2026, and currency pairs and tickers you lock into a portfolio are exempt.
An error in a backtest where a decision uses information that was not yet public on that date, for example a company's annual results used before they were published. Using point-in-time data and walk-forward testing reduces it.
M
The part of a broker account that Acubic manages, defined by the capital figure you assign when you connect. Rebalance orders are sized from that figure, and holdings outside it are never touched.
A company's size measured by its share price times the number of shares outstanding. Acubic can rank assets by market capitalization and compares companies in US dollars.
An allocation that gives larger companies larger weights, in proportion to their market capitalization, usually with a maximum per asset. Acubic offers market-cap weighted allocations capped at 10% or 20% per asset.
The classical method that chooses weights to balance expected return against variance, using past returns as inputs. It is the reference point other methods are judged against, and it can concentrate heavily when the return estimates are noisy.
The tendency of assets that have risen over a recent window to be ranked above those that have not. Acubic offers a twelve-month momentum ranking; it ranks past price changes and does not predict future ones.
O
The default Acubic automation mode. Each scheduled rebalance prepares the orders and waits, and nothing is sent to the broker until you approve it.
Evaluating a method on data that was not used to design or tune it. A result that only holds in-sample, on the data it was fitted to, is much weaker evidence than one that survives out-of-sample.
Fitting a method so closely to past data that it captures noise instead of a repeatable pattern. Testing many configurations against the same history raises the chance that a strong result reflects selection rather than signal.
P
Data recorded as it was known on each historical date, not as later revised. Acubic uses a company's financial statements only from the date they became publicly knowable.
The gap that opens between target weights and actual weights because holdings rise and fall by different amounts. Drift changes a portfolio's risk, not only its percentages, which is why portfolios are rebalanced.
R
An allocation that gives the best-ranked assets the largest weights, stepping down evenly through the ranking. Acubic also offers blends of rank weighting with equal weight or market-cap weighting.
How often a portfolio is brought back to target. In Acubic it is a calendar setting on the broker connection (monthly, quarterly or annually), not a drift threshold, so drift between two dates stays until the next scheduled run.
Trading a portfolio back toward its target weights by trimming what has grown and adding to what has shrunk. A rebalance trades only the difference between current and target holdings, not the whole portfolio.
The return on an asset treated as having no default risk, usually short-term government bills. Risk-adjusted measures such as the Sharpe ratio subtract it from a portfolio's return before comparing it with risk.
S
Return above the risk-free rate divided by volatility: how much return an investment produced per unit of total risk. It is a description of the past, and a high value over one period does not carry forward.
The difference between the price you expected for an order and the price it actually filled at, caused by spreads, timing and thin liquidity. Backtests that fill at historical closing prices cannot show it fully.
A variant of the Sharpe ratio that divides excess return by downside volatility only, so upward swings are not counted as risk. Acubic can use it as a ranking factor.
The distortion that appears when a backtest uses only assets that still exist today, leaving out those that were delisted or acquired. Their absence flatters historical results, and a universe built from current listings carries it.
T
The share of the portfolio an asset is supposed to hold, as a percentage of total capital. Target weights are what a rebalance compares your live holdings against.
How much of a portfolio is traded over a period. Higher turnover means more transaction costs and, in taxable accounts, potentially more tax events.
U
A European regulatory framework for funds that can be marketed to retail investors across the EU. European brokers typically list UCITS ETFs rather than US-listed ones, so a UCITS fund tracking the same index is the usual like-for-like equivalent.
V
How widely returns swing around their average, usually the annualized standard deviation of returns. Higher volatility means a wider range of outcomes in both directions.
W
A backtest that, at each rebalance date, rebuilds the portfolio using only data available up to that date, then holds it forward until the next date. It reduces look-ahead bias, but it does not remove the effect of searching the same history many times.
These definitions are educational and describe how a method works, not how an investment will perform. For the data Acubic uses and where its models fall short, read the methodology. To see drift in numbers, try the rebalancing calculator.