What is Fair Value Modelling?
The share price and US dollar foreign exchange on-line products both provide users with estimates of the fair value of share prices and exchange rates respectively. The fair value of a share price or exchange rate is the share price or exchange rate value that is expected on the basis of a set of driver variables. Multivariate regression modelling has been used to identify these driver variables as statistically significant over a particular historical period of time. The actual share price or exchange rate can rise above or fall below the fair value for a number of reasons. The modelling demonstrates, however, that the actual share price or exchange rate always gravitates back to the fair value.
This equation is estimated on the basis of historical data. By substituting values for the driver variables into this equation we are able to obtain the value of the share price or exchange rate that is expected on the basis of the driver variables.
The multivariate regression equations are estimated on the basis of daily historical data. The daily driver variables that are used in order to explain the movements in the share price and exchange rate are proxies for :
- Movement in the market.
- The strength of the underlying economy and core inflationary pressures.
- Short term wholesale costs faced by financial institutions.
- The degree of risk aversion in financial markets.
- The impact higher unit fuel costs have on certain segments of the market.
- The strength of global commodities and energy demand.
- The impact of that exchange rate on the price of US dollar denominated imported goods.
- News in financial and economic markets that transmit information via the Australian dollar/US dollar exchange rate.
- The impact of other exchange rates on the price of imported goods denominated in other currencies.
- News in financial and economic markets that transmit information via exchange rates.
FX Fair Value Analysis
In addition to the variables presented above the FX fair value modelling also requires official interest rates as set by Central banks around the world.
The monthly statistical variance of the Dow Jones Industrial Average stock index as a proxy for volatility and risk in financial markets.
Interval Estimates for Future Fair Value
We estimate a range of fair values in the case of future estimates of the fair value (low, point estimate and high). This range of values takes into consideration the statistical difficulty in estimating future values. Notably it takes into consideration the inability of the multivariate regression equation to fully account for all of the actual movements in share prices and the exchange rates that are observed historically (also known as sampling error). It does not take into consideration the volatility of individual slope parameters that are estimated as part of the multivariate regression analysis.
Impact of Other Variables
The fair value represents the share price or foreign exchange value that is expected on the basis of the driver variables. From time to time there will be movements in other variables that will affect the share price, particularly variables that are peculiar to particular companies, or markets in which particular companies operate. Users will need to consider whether these other variables are driving the share price above or below fair value.
Examples of these other variables may include, but is not limited to, volatility, disruptions to or delays in the supply-side of various markets (for example airline capacity in the case of an analysis of airline share prices, or coal supply in the case of an analysis of coal mining company shares), actual or the threat of industrial action, weather conditions, natural disaster or other acts of nature (floods, volcanic eruptions, earthquake, drought), war, civil unrest and expected or actual Government regulation.
As indicated above, the actual share price or foreign exchange value may rise above or fall below the fair value. This may be because of random sampling error or because of the impact of non-driver variables as explained above. The number of days over which the actual share price or foreign exchange rate rises above or falls below the fair value is called a run.
Historical analysis can be conducted on these runs. Specifically, it is possible to determine the relative frequency of different lengths of runs. For example, it is possible to determine the percentage of days over the historical period examined in which the actual share price or foreign exchange value exceeded the fair value for just one day.
In the on-line share price and foreign exchange fair value models to follow an important part of those products will be a historical analysis of runs.
How often is the Fair Value Model re-estimated?
The fair value model equation is re-estimated on a weekly basis.
Up to five new daily sets of data will be added to the share price and foreign exchanged history and the multivariate regression model will be re-estimated.
Disclaimer and Limitations of the Model
The information contained in this website is provided ‘as is’ without warranty of any kind. The entire risk as to the results and the performance of the information obtained from this website is assumed by the user, and in no event shall Webber Quantitative Consulting Pty Ltd be liable for any consequential, incidental or direct damages suffered in the course of using the information in this website.