VaR and its use to measure risk TAdvisorpedia 27 May 2019 , No hay comentarios When talking about investments portfolio, immediately we come across with the concept of Value at Risk (VaR). A specific indicator with different calculation methods to analyse risk. Because of its importance to the portfolio investment methodology, we have made a strong analysis and comments about the main advantages and disadvantages. The concept of Value at Risk (VaR) JP Morgan in the U.S. was the first to come up with the concept of Value At Risk at the end of the 80s. But how did it started? The Weatherstone’s president wanted to explain in one page the risk involved in the risk of losing the value of a portfolio. With this in mind, he created a new indicator that could summarize and evaluate the risk of an investment portfolio. Let’s look at an example to really understand it. If we set up a portfolio X with a daily VaR of 20,000 euros at a 99% confidence level. This actually means that in the next 24 hours, the most that this portfolio will lose with a 99% confidence level is 20,000 euros and chances of losing more than those 20,000 euros is 1 every 100 sessions. Methods of calculating VaR Variance-Covariance method: This method is based on historical data to forecast market volatility and correlations. The result is used with a statistical model in the portfolio assuming a normal asset distribution. Historical method: The main difference between the variance-covariance and historical method lays on the distributions used, as the historical method does not assume a normal distribution. This method gives a profit-loss analysis assuming the historical risk distribution of the assets in the portfolio and with it, we can get a daily forecast for the profit-loss, distribution and main distribution outliers for a more accurate VaR measurement. Montecarlo simulation: This method uses random values to simulate the prices of the asset over a specific time horizon. This model has proven to make produce more accurate results, but taking as a basic component, assets that do not have historical information nor a normal distribution Advantages and Disadvantages We cannot truly say which estimation method is the best as each one of them has different objectives and purposes. But we have created a small guideline of the benefits and application of each method: The variance-covariance method is mainly used when an investment portfolio does not have financial derivatives. On the other hand, it shows some disadvantages when it comes to positions for non-linear assets or with a very unusual distribution. The historical method has a certain advantage in which the latest risk value is more significant to the actual risk value. Also gives a good estimation for non-linear assets. On the other hand, when the past is not significant to the future distribution it might not be recommended to use this method. The Montecarlo method is also recommended when a portfolio has non-linear assets. However, when the investment portfolio has a large asset composition or with financial derivatives is not recommended because might use a lot of computing resources. Overall, the use of any of these models is based on the characteristics of the assets, as there are a lot of estimation methods according to the VaR methodology. which makes it difficult to choose one specific method. Turning this into a decision point for every financial institution to satisfied their goals and needs. Never the less, there is additional analysis that can be done such as Stress testing, Backtesting or the Expected Shortfall. The Value at Risk (VaR) is one of the most common risk assessment indicators at a compliance level.