Викторина, связанная с руководством · Средний Уровень

Прогнозирование волатильности с помощью викторины по машинному обучению

Тестирует волатильность в зависимости от направления, базовый уровень HAR, дизайн вне выборки, утечку, метрики оценки и пределы прогноза.

Предварительный просмотр каждого вопроса8 вопросы

Ответы и пояснения появляются по мере прохождения теста.

  1. What does a volatility forecast estimate?
    • Expected return over the forecast horizon
    • Probability of a positive return
    • Future variability of returns over a stated horizon
    • Expected maximum portfolio drawdown
  2. Как HAR-RV обычно используется в исследованиях волатильности?
    • A GARCH-family model based on conditional variance
    • A benchmark combining realized volatility over daily, weekly, and monthly horizons
    • An option-implied volatility index
    • A rolling standard deviation over one fixed window
  3. Why use out-of-sample evaluation?
    • To test forecasts on information not used to fit the model
    • To estimate fit after selecting parameters on the entire sample
    • To choose the most favorable test period and report that result
    • To refit on the holdout and call the fit independent
  4. Which input would create look-ahead leakage in a forecast made at the end of today?
    • Using a trailing volatility value calculated only from earlier prices
    • Using yesterday’s return, which was already available at forecast time
    • Fitting a scaler only on earlier training observations
    • Using information that was unavailable at the forecast date
  5. Does a lower forecast error prove trading profitability?
    • Yes; a lower out-of-sample error proves positive net returns
    • No; volatility forecasts cannot inform position sizing when expected returns are unchanged
    • No; costs, execution, and portfolio rules also matter
    • Yes; accurate volatility forecasts determine profitable trade timing
  6. Зачем сравнивать с простыми линейными базовыми линиями?
    • To test whether complexity adds value beyond a credible benchmark
    • Because linear models always capture every volatility pattern
    • Because a baseline removes the need for out-of-sample testing
    • Because an in-sample gain guarantees better forecasts
  7. Какой вопрос касается направления возврата, а не величины волатильности?
    • It predicts whether the asset price will rise
    • It estimates variation in returns over the stated horizon
    • It compares variability estimates across horizons
    • It reports uncertainty around a variability estimate
  8. Какой отчет предоставляет контекст, необходимый для воспроизведения и интерпретации прогноза волатильности?
    • Target, horizon, and assets, without costs, uncertainty, or evaluation design
    • Target, horizon, assets, fees, uncertainty, and test design
    • Assets and selected model, without fees or evaluation design
    • A score and chart, without target, horizon, or data-split details