FN3142 Quantitative Finance
Summer 2021
Question 1
Denote the price process of Bitcoin by P , and consider (T + 1) consecutive end-of-year price observations denoted by P0 , P1, …, PT . Denote the simple net return series by and the logarithmic return series by rt ā” log Pt – log Pt-1 for t = 1, …, T. Assume that returns are independently and identically distributed over time.
(a) [15 marks] Under the assumption that Bitcoinās annual logarithmic returns are normally distributed with a mean of μ and a variance of Ļ 2 , derive a formula for the diference between the log expected simple gross return and the expected log return:
log (E [1 + Rt]) – E [rt]
(b) Consider the standard average estimator of the mean rate of return. Show that this estimator is unbiased in the sense that E[μ-] = μ, and derive its standard deviation, Ļ[μ-], as a function of the model parameters.
(c) [15 marks] Assume instead that you observe Bitcoin prices more frequently, in particular N times per year (at equal distance from each other) for T years; for example, monthly ob- servations would mean N = 12. Consider now the simple average estimator of the annualised log return μ based on these N Ć T return observations, and let us denote it by Ė(μ) . Show that the standard deviation of this estimator, Ļ[Ė(μ)], does not depend on the frequency N.
(d) [20 marks] Suppose that a researcher tells you that for a given choice of N and T she estimated the annualised mean Bitcoin return to be Ė(μ) = 10%. Suppose also that the annualised volatility of Bitcoin log returns is known to be Ļ = 45%. What choice of T would imply that her estimate has a standard deviation of Ļ[Ė(μ)] = 1%, i.e., one-tenth the size of the mean estimate? Discuss whether it is possible to have conļ¬dence in the rate of return of Bitcoin with reasonable accuracy given the available historical data.
Let us now assume that annual log Bitcoin returns have a non-zero autocorrelation at lag 1 denoted by parameter Ļ, but beyond lag 1 there is zero autocorrelation.
(e) [15 marks] Derive E[μ] and Ļ[μ-] under this assumption, where μ refers to the estimator in part (b).
(f) [15 marks] Discuss how a non-zero Ļ afects your conclusion for part (d).
Consider the following AR(2) process:
zt = Q0 + Q1 zt-1 + Q2 zt-2 + εt , (1)
where εt isa zero-mean white noise process with variance Ļ2 , and assume j Q1 j , j Q2 j , j Q1 +Q2 j < 1, which together ensure zt is covariance stationary.
(a) [15 marks] Calculate the conditional and unconditional means of zt , that is, Et-1 [zt] and E [zt].
(b) [15 marks] Let us now set Q2 = 0. Calculate the conditional and unconditional variances of zt , that is, V art-1 [zt] and Var [zt].
(c) [20 marks] Keeping Q2 = 0, derive the autocovariance and autocorrelation functions of this process for all lags as functions of the parameters Q1 and Ļ .
Suppose now that Q2 ā 0, and let us denote the autocovariance at lag k by āk = Cov[zt , zt-k ].
(d) [15 marks] Using equation (1), write down a recursive formula for āk , i.e., express āk as a function of āk -1 , āk -2 and the model parameters.
(e) [20 marks] Apply this recursive formula for k = 1 and k = 0, and explain how to solve for the whole autocovariance function {āk }kā„0 . Note: No need to derive the exact values!
Hint: think about what ā-1 and ā-2 mean?
(f) [15 marks] Can a linear transformation of the zt process be represented by an AR(1) process? That is, do appropriate h, Ī“0 , and Ī“1 constants exist such that the process deļ¬ned as yt = zt + hzt-1 satisļ¬es
yt = Γ0 + Γ1yt-1 + Et ,
where Et is a zero-mean white noise process? Explain.
Question 3
Consider two unbiased forecasts f1,t and f2,t of the mean-zero quantity yt+h. Denote the indi-vidual errors by ei,t+h yt+h -fi,t , i = 1, 2, the mean-square forecast errors by Ļi(2) E[ei(2),t+h] for i = 1, 2, and the correlation between the individual errors by Ļ corr[e1,t+h, e2,t+h].
Consider the linear forecast combination
fc,t ā” Ī»f1,t + (1 – Ī»)f2,t where Ī» ā [0, 1],
and denote the error associated with the combination by ec,t+h yt+h – fc,t.
(a) [20 marks] Compute the mean forecast error, E[ec,t+h], and the mean-squared forecast error, E[ec(2),t+h], of the combination forecast.
(b) [20 marks] By minimising the mean-square-error loss of the combination show that the optimal weights are functions of the accuracy of each prediction and the correlation parameter, and given by
(c) [20 marks] Assuming the forecasts are equally accurate, i.e., Ļ 1 = Ļ2 = Ļ, derive the optimal weights and the mean-squared forecast error of the combined forecast.
(d) [20 marks] Compute the range of the correlation parameter Ļ for which the optimal combined forecast displays diversiļ¬cation beneļ¬ts, that is, its mean-squared forecast error, E[ec(2),t+h], is lower than the mean-squared errors of the original forecasts, Ļ 1(2) and Ļ2(2). Explain your results.
(e) [20 marks] Imagine that you have generated a forecast with the optimal weights calculated above, but then it turns out that the projection has poor performance in the sense of having a low R2 in a regression of the forecasted variable on a constant and the forecast. What would be the properties of a data-generating process that would lead to such a result?
Question 4
Assume that daily returns are conditionally normally distributed and given by
rt+1 = μ + Ļt+1vt+1 , vt+1 i~.i.d N (0, 1),
where the time increment between t and t + 1 is one day, μ is constant, and Ļt+1 denotes an estimate as of time t for the conditional standard deviation one day ahead.
(a) [20 marks] The 1-day Value-at-Risk at the critical level Q, is defined as
Show that the exact formula for VaR at the Q critical level and 1-day horizon is given by
where Φ is the standard normal cumulative density function.
Note: The deļ¬nition in equation (3) follows the notation of the subject guide. Using an alternative deļ¬nition based on losses leads to a diferent VaR formula, which is also accepted if correct.
(b) [30 marks] Describe the historical simulation, the RiskMetrics (also known as exponen- tially weighted moving average), and the GARCH normal approaches to measuring Value- at-Risk in your own words. Discuss their beneļ¬ts and shortcomings.
(c) [25 marks] The expected shortfall ESt
α
+1 at the critical level α and 1-day horizon can be deļ¬ned as
Using the VaR formula from part (a), derive the following formula for the 1-day expected shortfall at critical level Q:
where Ļ is the standard normal probability density function.
Hint: If z follows a standard normal distribution, z ~ N (0, 1), we have
(d) [25 marks] Discuss the beneļ¬ts and shortcomings of using expected shortfall as a risk measure instead of Value-at-Risk.

![[SOLVED] FN3142 Quantitative Finance Summer 2021 Python](https://assignmentchef.com/wp-content/uploads/2022/08/downloadzip.jpg)

![[Solved] Program that has three functions: sepia(), remove_all_red(), and gray_scale()](https://assignmentchef.com/wp-content/uploads/2022/08/downloadzip-1200x1200.jpg)
Reviews
There are no reviews yet.