The AVC Backtest: Establishing Data Validity
The Asset Valuation Curves (AVC) provides clients with the information needed to make informed decisions and manage risk in used equipment assets. Meanwhile, the AVC Backtest functions as an addendum to the primary AVC releases and aims to ascertain the validity and soundness of the underlying data that makes up the AVC.
Our Backtests utilizes a mix of data points in order to perform this validation check. These data points stem from mainly two sources: the underlying AVC data as well as sales data from associated brands that are part of the RB Global umbrella. We mainly import data from Ritchie Bros. Auctioneers and Mascus, which helps us to cover both auction and online listings data, respectively.
Both sets of data points undergo an extensive cleaning process before they are fit for usage in the backtesting process. However, it is not only the underlying data that has being examined closely. The methodology of the backtest has also been subjected to intense scrutiny in order to ensure that it aligns accordingly to the European Central Bank’s Capital Adequacy Directive, Basel III.
The Basis of a Backtest
Our AVC Backtests comes in the form of an in-depth report. In it we list all the steps of the backtesting process in greater detail, covering a wide range of different aspects. These aspects have been further arranged according to the following key sections: the methodology of the backtesting process as well as the results of the performed process.
In short, the methodology section takes a closer look at how the underlying data points are collected, altered, and utilized throughout the backtest. It also discusses some of the choices made regarding the presentation of said data points. Our primary choice of presentation settled on charts and curves. Each chart covers one respective product group that is within the requested scope of the backtest. Similarly to the AVC releases, these charts help to display how the prices of objects present within said product groups depreciate over time.
However, the charts of the backtest also include other pricing aspects derived from the imported sales data. This allows for direct comparisons to take place between the values of the AVC releases and the real-life sales data; thus showing how well the AVC performs in real-life scenarios. This is the second key aspect of the report, namely the presentation of the results of the backtesting process.
A third, shorter section covering regional market trends is also included in the report. This allows us to better contextualize the findings of the backtest and put the results into a more regional perspective. All in all, these steps combined help to show how the underlying AVC data remains valid and sound over a multitude of different product groups.
The Building Blocks of a Backtest Chart
We have chosen to present our findings in the form of charts, as can be seen in the example to the right. Each chart represent a specific product group and covers how the prices of the affected equipment depreciate over a ten-year-span. Five curves in total are used in each chart, each of which cover their own distinct pricing aspect. Including all five of these curves in the same chart allows for us to paint a more nuanced picture over how the price depreciates over time. This method also allows us to show how the different pricing aspects relate to each other in a simple yet effective manner.
The cornerstone of each chart is undoubtedly the curve representing the AVC Index values. This curve presents the underlying data behind the AVC expressed as a percentage of the initial equipment value. The remaining four curves are all based on statistical measures derived from our imported sales data and are thus meant to be directly compared to the AVC Index curve and its values. The Average Price curve represents, just as the name implies, the average price of sold objects present within the chosen product group. Meanwhile, the P25, P50, and P75 curves all represent specific price percentiles of sold objects. Each curve covers the 25th, 50th, and 75th pricing percentiles, respectively, and are included in the chart to better show the price range between expensive and inexpensive objects.