What does it mean for two non-stationary time series to be cointegrated?
Their individual levels are non-stationary, but a linear combination of them is stationary; they share a stable long-run relationship.
2
New cards
In the Engle–Granger approach, what is tested after regressing one time series on another?
The regression residuals are tested for a unit root. Evidence that the residuals are stationary supports cointegration.
3
New cards
What does failure to reject a unit-root null for the cointegrating regression residuals imply?
The residuals appear non-stationary, so the evidence does not support cointegration.
4
New cards
Why is it not enough for both series to be individually stationary to establish cointegration?
Cointegration is a concept for non-stationary series that share a stationary long-run combination; stationary series do not meet that defining setup.
5
New cards
What is the role of an error-correction model when variables are cointegrated?
It captures both short-run changes and the adjustment back toward the long-run equilibrium relationship.
6
New cards
What does the mean-reverting level of an AR(1) process represent?
The long-run level toward which the process tends to revert when it is mean-reverting.
7
New cards
What does significant serial correlation in regression residuals indicate?
Residuals are related across time, suggesting the model has not captured all relevant time-series dynamics.
8
New cards
What is a common response when regression residuals show significant serial correlation?
Re-specify the model to capture the omitted dynamics, such as by adding an appropriate lag.
9
New cards
What is autoregressive conditional heteroskedasticity (ARCH)?
A pattern in which the conditional variance of errors depends on past squared errors; volatility clusters over time.
10
New cards
What does an ARCH finding suggest about a model's residuals?
Their variance is time-varying and predictable from past shocks, so the constant-variance assumption is not appropriate.
11
New cards
What is conditional heteroskedasticity?
Error variance changes systematically with one or more explanatory variables or other conditioning information.
12
New cards
How does unconditional heteroskedasticity differ from conditional heteroskedasticity?
Unconditional heteroskedasticity is not related to the regression's explanatory variables; conditional heteroskedasticity is related to them.
13
New cards
What pattern in a residual-versus-predictor plot suggests conditional heteroskedasticity?
The spread of residuals changes as the predictor changes, such as a fan-shaped pattern.
14
New cards
Why can heteroskedasticity undermine conventional regression inference?
It can make the usual standard errors and significance tests unreliable, even when coefficient estimates remain unbiased under other assumptions.
15
New cards
What does a statistically insignificant residual-autocorrelation slope suggest?
There is insufficient evidence of serial correlation in the residuals under the test being used.
16
New cards
What is term frequency (TF) in text analytics?
A measure of how often a token or term appears in a document or collection of documents.
17
New cards
Why can extremely rare tokens be problematic features in a text-classification model?
They may occur in only a few cases and cause the model to associate an unusual token with a class, encouraging overfitting.
18
New cards
Why can extremely common tokens be problematic features?
They may carry little distinguishing information because they appear across many documents.
19
New cards
What is vocabulary pruning?
Removing unhelpful tokens from the feature set, often including terms that are extremely frequent or extremely sparse.
20
New cards
What is overfitting?
A model learns noise or idiosyncrasies in the training data and performs poorly on unseen data.
21
New cards
Why can an unconstrained decision tree overfit?
It can keep splitting until it learns the training observations very closely, including noise.
22
New cards
How can decision-tree overfitting be reduced?
Use constraints or regularization, such as limiting tree depth, requiring minimum observations in leaves, or pruning.
23
New cards
Why can a nonlinear support vector machine be prone to overfitting?
Flexible nonlinear boundaries can fit training data very closely and become unnecessarily complex.
24
New cards
What is the purpose of regularization in machine learning?
To penalize model complexity and reduce overfitting, improving generalisation to unseen data.
25
New cards
What happens to ordinary in-sample R-squared when additional explanatory variables are added to a regression?
It can increase or stay the same; it cannot decrease simply because another regressor was added.
26
New cards
What is precision in a classification model?
The proportion of predicted positive cases that are actually positive.
27
New cards
What is recall in a classification model?
The proportion of all actual positive cases that the model correctly identifies.
28
New cards
When should a model prioritise recall?
When missing actual positive cases (false negatives) is especially costly.
29
New cards
What is a false negative?
An actual positive case that the model incorrectly classifies as negative.
30
New cards
What is a false positive?
An actual negative case that the model incorrectly classifies as positive.
31
New cards
What does logistic regression model in a binary classification setting?
The probability of an outcome, commonly through a transformation of a linear combination of explanatory variables.
32
New cards
Why is the average predicted probability generally not the same as the probability calculated using average inputs?
The logistic transformation is nonlinear, so applying it to average inputs need not equal averaging individual predicted probabilities.
33
New cards
What is survivorship bias in investment backtesting?
Using only entities that survived to the end of the sample, which excludes failed, delisted or removed entities and can make historical performance look too favourable.
34
New cards
How can point-in-time constituent data reduce survivorship bias?
At each historical date, use only the securities that actually belonged to the index or universe at that date.
35
New cards
What is look-ahead bias?
Using information in a historical simulation that would not yet have been available at the time of the simulated investment decision.
36
New cards
Why do monthly rolling-window backtests often have overlapping in-sample periods?
Each new monthly window shifts forward only one month, so most of its historical estimation period overlaps the previous window.
37
New cards
What is a key limitation of VaR during periods of market stress?
Correlation risk: relationships between assets can change sharply, with correlations often rising and diversification benefits weakening.
38
New cards
Why might relative VaR add little information for an absolute-return strategy benchmarked against cash?
The strategy's relative risk to cash may be effectively the same as its absolute risk, so it may duplicate information already provided by VaR.
39
New cards
For what type of investor is surplus-at-risk particularly relevant?
Pension funds, because it focuses on the risk that assets will be insufficient relative to liabilities.
40
New cards
What is a currency board?
A strict fixed-exchange-rate arrangement in which the monetary authority commits to exchange domestic currency for a reserve currency at a fixed rate and backs the monetary base with foreign reserves.
41
New cards
How does a currency board differ from a conventional fixed peg?
A currency board imposes tighter institutional and reserve-backing constraints, leaving less scope for discretionary monetary policy.
42
New cards
What is a crawling peg?
An exchange-rate regime in which the currency is adjusted gradually, often according to a preannounced rate or set of indicators.
43
New cards
What is a managed float?
An exchange rate that is primarily market-determined but in which the central bank may intervene to influence its path.
44
New cards
What is the main monetary-policy trade-off under a hard currency peg?
The country gives up much of its independent monetary-policy flexibility to maintain the exchange-rate commitment.
45
New cards
What is a tariff?
A tax imposed on imported goods, generally raising their domestic price and protecting domestic producers.
46
New cards
What is an import quota?
A direct limit on the quantity of a good that may be imported.
47
New cards
How does a quota differ from a tariff?
A tariff raises the cost of imports through a tax; a quota restricts the quantity imported.
48
New cards
What is a voluntary export restraint?
A limit on exports that the exporting country agrees to impose, usually following pressure from the importing country.
49
New cards
What is trade diversion in a customs union?
Trade shifts from a lower-cost non-member producer to a higher-cost member producer because the non-member faces the union's external tariff.
50
New cards
What is trade creation in a customs union?
Higher-cost domestic production is replaced by lower-cost imports from another member, improving resource allocation.
51
New cards
What is the difference between extensive and intensive economic growth?
Extensive growth comes from using more inputs; intensive growth comes from using inputs more productively.
52
New cards
What is total factor productivity (TFP)?
The portion of output growth not explained by growth in measured labour and capital inputs; it reflects improvements in overall productive efficiency and technology.
53
New cards
What is the transmission mechanism of monetary policy?
The channels through which policy actions affect financial conditions, aggregate demand, output and inflation.
54
New cards
What is the liquidity trap concept?
A situation in which interest rates are very low and conventional monetary easing becomes less effective because people are willing to hold additional money rather than spend or invest it.
55
New cards
What is the trade-off theory of capital structure?
Firms balance the tax benefits of debt against the expected costs of financial distress and other costs associated with leverage.
56
New cards
What is the pecking-order theory?
Firms prefer internal financing first, then debt, and issue equity as a last resort because of information asymmetry and issuance costs.
57
New cards
Why might managers prefer internal financing under the pecking-order theory?
It avoids external issuance costs and the adverse signals or valuation concerns that can accompany issuing securities.
58
New cards
What is the agency cost of debt?
Conflicts between shareholders and debtholders that can encourage risk shifting, underinvestment or other actions that transfer value from creditors.
59
New cards
What is the agency cost of equity?
Costs arising from conflicts between managers and shareholders, including managerial actions that do not maximise shareholder value.
60
New cards
What is a share repurchase?
A company buys back its own shares, reducing shares outstanding and returning cash to shareholders who sell.
61
New cards
How does a cash dividend differ from a share repurchase?
A dividend distributes cash to shareholders proportionally; a repurchase pays cash to shareholders who tender or sell shares.
62
New cards
What is a stock dividend?
A distribution of additional shares to existing shareholders rather than cash; it increases the number of shares held but does not itself transfer company value.
63
New cards
What is the principal–agent problem in corporate governance?
Managers (agents) may pursue their own interests rather than the interests of shareholders (principals).
64
New cards
How can board independence help mitigate agency conflicts?
Independent directors can provide oversight and challenge management decisions on behalf of shareholders.
65
New cards
What is the difference between a one-tier and two-tier board structure?
A one-tier structure combines executive and non-executive directors on one board; a two-tier structure separates management and supervisory boards.
66
New cards
What is a related-party transaction?
A transaction between a company and a party with a relationship to the company, such as a director, executive or significant shareholder; it can create conflicts of interest.
67
New cards
What does the cash conversion cycle describe?
The time between paying suppliers for inventory and collecting cash from customers, reflecting how efficiently working capital is managed.
68
New cards
What is a divestiture?
The sale or disposal of a business unit, subsidiary, asset group or investment by a company.
69
New cards
What is a spin-off?
A transaction in which a parent distributes shares of a subsidiary to its existing shareholders, creating a separately traded company.
70
New cards
What is a primary market?
The market in which newly issued securities are sold and the issuer raises capital.
71
New cards
What is a secondary market?
The market in which existing securities trade between investors; the issuer generally does not receive the proceeds of these trades.
72
New cards
What is a dealer market?
A market in which dealers quote bid and ask prices and trade from their own inventory or risk positions.
73
New cards
What is an order-driven market?
A market in which buy and sell orders are matched, typically according to price and time priority.
74
New cards
What is a dark pool?
A private trading venue where orders are not publicly displayed before execution, often used for institutional trading.
75
New cards
What is the bid–ask spread?
The difference between the price at which a dealer or market will buy and the price at which it will sell; it is a component of trading cost.
76
New cards
What is market impact?
The price movement caused by a trade itself, often because the order consumes available liquidity or reveals information.
77
New cards
What is the difference between enterprise value and equity value?
Enterprise value represents the value of the operating business to all capital providers; equity value is the residual value attributable to common shareholders.
78
New cards
Why is enterprise value often paired with an operating earnings measure?
Enterprise value reflects claims of both debt and equity providers, so it is compared with earnings available before financing costs, such as EBITDA.
79
New cards
What is a justified price-to-book ratio conceptually linked to?
A firm's expected profitability relative to its cost of equity and its growth prospects; stronger sustainable returns on equity can support a higher price-to-book ratio.
80
New cards
What is residual income in equity valuation?
The earnings remaining after charging equity capital for its required return; value is created when earnings exceed that equity charge.
81
New cards
What is the clean-surplus relation?
The accounting relationship in which changes in book value are explained by earnings and distributions to shareholders, with other comprehensive income adjustments treated consistently.
82
New cards
What is a key limitation of using P/E ratios to compare companies?
Differences in accounting policies, earnings quality, capital structure, growth, risk and cyclicality can make apparently similar P/E ratios non-comparable.
83
New cards
Why can a negative-earnings company not be meaningfully compared using a conventional P/E ratio?
The ratio becomes economically difficult to interpret when earnings are negative.
84
New cards
What is a sum-of-the-parts valuation?
A valuation that estimates the value of separate business segments individually and then combines them, with appropriate adjustments for corporate items and claims.
85
New cards
What is a control premium?
The additional amount a buyer may pay to obtain control of a company beyond the value implied by a minority shareholding.
86
New cards
What is a minority discount?
A reduction in value reflecting the limited influence and control rights of a non-controlling ownership interest.
87
New cards
What is a rights offering?
An issuance in which existing shareholders are given rights to purchase additional shares, often in proportion to their current holdings.
88
New cards
What is a tender offer?
A public offer to shareholders to sell their shares at specified terms, commonly used in acquisitions or share repurchases.
89
New cards
What does the weak form of market efficiency assert?
Prices reflect information contained in past market data, such as historical prices and trading volume.
90
New cards
What does the semi-strong form of market efficiency assert?
Prices reflect all publicly available information.
91
New cards
What does the strong form of market efficiency assert?
Prices reflect all information, both public and private; even insiders cannot consistently earn abnormal returns.
92
New cards
What is an anomaly in the context of market efficiency?
A persistent return pattern that appears inconsistent with a market-efficiency model or asset-pricing benchmark.
93
New cards
What is the difference between an active and passive equity strategy?
An active strategy seeks to outperform a benchmark through security selection or market timing; a passive strategy seeks to track a benchmark.
94
New cards
What is systematic risk?
Risk arising from broad market or economic factors that cannot be eliminated through diversification.
95
New cards
What is unsystematic risk?
Company- or industry-specific risk that can be reduced through diversification.
96
New cards
Why does diversification reduce unsystematic risk?
Holding securities with different company-specific drivers reduces the impact of any single issuer or event on the overall portfolio.
97
New cards
What is an equity market-neutral hedge fund strategy?
A strategy that seeks to profit from security selection while keeping broad market exposure relatively neutral through offsetting long and short positions.
98
New cards
What is typically the main source of skill in an equity market-neutral strategy?
Security selection; market timing is generally secondary.
99
New cards
Why may equity market-neutral managers use relatively high leverage?
Because many broad market and sector risks are hedged, managers may use leverage to pursue meaningful return targets, although leverage still magnifies losses and other risks.
100
New cards
What is an event-driven hedge fund strategy?
A strategy that seeks to profit from corporate events such as mergers, restructurings, bankruptcies or distressed situations.