Chapter 8 : Perform the Analysis -Predictive Analytics

These flashcards cover key concepts from the Predictive Analytics chapter, including definitions, distinctions, techniques, and examples related to predictive analysis and its applications.

Predictive Analytics

Analyzes historical data to predict future outcomes, likelihoods, or behaviors, providing foresight for proactive decision-making.

Deterministic Analysis

Provides certain outcomes based on known facts and established data, offering definite results without uncertainty.

Probabilistic Analysis

Focuses on the likelihood and probability of future events, using statistical models to quantify risk and potential outcomes.

Probabilistic Model Example

Predicting loan repayment risk; uses financial and behavioral data (e.g., credit scores) to calculate repayment probability.

Classification

A supervised learning technique in predictive analytics to categorize data into predefined groups (e.g., high-risk/low-risk, fraudulent/legitimate).

Altman's Z-score

Formula predicting bankruptcy likelihood (within 2 years) using five financial ratios: $(Working Capital) / (Total Assets)$, $(Retained Earnings) / (Total Assets)$, $(Earnings Before Interest and Taxes) / (Total Assets)$, $(Market Value of Stockholders’ Equity) / (Book Value of Total Debt)$, and $(Sales) / (Total Assets)$.

Z-score < $1.80$

Indicates a company is in the 'distress zone,' facing significant risk of bankruptcy or financial insolvency.

Base Rate Fallacy

Cognitive bias where predictions ignore historical averages (base rate) and over-rely on new, specific information, leading to inaccurate conclusions.

Regression Analysis

Statistical technique in predictive analytics modeling the relationship between a dependent variable and independent variables to estimate or predict outcomes.

Time Series Analysis

Statistical method using historical data patterns (trends, seasonality) to predict future values of a variable, common in accounting and finance.

Financial Statement Persistence

The continuity and durability of financial statement variables (e.g., earnings) over time, indicating how reliably current performance predicts future performance.

Hypothesis Testing

Evaluates relationships between variables and impact of decisions in predictive analytics. Validates assumptions, determines statistical significance, enhancing model reliability.

Beneish M-Score Factors

Factors predicting financial statement fraud: increased Days Sales Receivable Index ($DSRI$), declined Gross Margin Index ($GMI$), declined Asset Quality Index ($AQI$), and increased Sales Growth Index ($SGI$).

Lender Loan Evaluation

Lenders assess creditworthiness for loan extensions by evaluating credit scores, employment history, existing debt levels, and debt-to-income ratios ($DTI$).

How is hypothesis testing used in accounting analytics?

To test statistical evidence for or against a business assumption

What is the expected relationship between requested loan amount and loan acceptance?

Negative

In Altman’s Z-score, what does X1 measure?

Liquidity

How can overfitting affect a predictive model in accounting?

It tailors the model too closely to historical data, reducing future accuracy

Which statistical technique is best suited for forecasting a continuous accounting value, such as sales?

Regression

What does RPA stand for?

Robotic Process Automation

How does employment length affect loan acceptance likelihood?

Longer employment increases acceptance

Which technique is used in forecasting quarterly sales?

Time Series analysis

What is a base rate fallacy in the context of accounting prediction?

Ignoring overall event frequency when interpreting specific results

What is the primary goal of predictive analytics in accounting?

To estimate future outcomes based on patterns in historical data

Which predictive technique categorizes observations into groups?

Classification

Base rates refer to...

Historical averages of event occurrence

Which of the following can be a dependent variable in regression?

Interest rate

What kind of outcome does regression produce?

Numeric Estimate

The Beneish M-score is designed to identify companies likely to be engaging in what?

Earnings Manipulation

Which model is used to estimate bankruptcy risk?

Altman’z Z-score

Altman’s Z-score is primarily used to predict which of the following?

Bankruptcy Risk

Which index is used to detect financial statement fraud?

Beneish M-score

What does a p-value below 0.05 typically indicate?

Strong evidence agai