Residual value forecasting of aircraft remains as much an art as a science but with each passing year, new datapoints emerge that should help to improve accuracy though the impact of any unexpected event is likely to remain underestimated.
The chief economist of the Bank of England recently stated that economic forecasts had been wrong before the financial crisis and wrong about the immediate impact of the Brexit vote to leave the European Union (EU). The effect of the collapse of Lehman Brothers in 2008 was wholly underestimated and the impact of the vote of Britain to leave the EU overestimated. In the latter case, the Bank of England correctly anticipated that there would be a significant weakening of the British pound but seriously underestimated consumer confidence such that today economic growth in the United Kingdom remains positive. Modelling the effects of shock events or “Black Swan” moments is inevitably subjective and prone to considerable errors. JK Galbraith once stated “The only function of economic forecasting is to make astrology look respectable” and seeking to forecast the future value of aircraft is just as difficult. Despite considerable depth and breadth of data, there is considerable variation in economic forecasts produced by governments around the world from one quarter to the next. The problem for both economist and the future value forecasting is that shock events, by their very nature, cannot be anticipated but what can be forecast is their impact on future demand.
While there are many thousands economists contributing to the wealth of data necessary to formulate economic forecasts upon which so many livelihoods depend – only to see a range of inaccuracy – there are perhaps less than 15 “Master Appraisers” in the world. These Master Appraisers have the experience and responsibility for creating, evolving and maintaining non-financial engineering and non-pure statistical based econometric future value models, though of course, many more certificated and non-certificated appraisers then rely on the forecasts created by such core Master Appraisers to prepare appraisals – without having had to create or refine their own model.
Over the last 50 years there have been approximately nine shock events that has caused a considerable variation between forecast values and the actual outcome – the spread of jets in the 1960s; the first oil crisis of the early 1970s; the Iran-Iraq war of the late 1970s; the collapse in oil prices in 1986; the first Gulf War in the early 1990s; the Asian crisis of the late 1990s; the events of 2001; bird flu and SARS; the financial crisis of 2008. There is therefore a need to assess the potential impact of shock events on aircraft values. Just as the science of meteorology can improve forecasting accuracy by examining past datapoints, then so too is future value forecasting able to refine existing models by incorporating the impact of past events.
As aircraft appraising has increasingly moved away from abstract base values towards more realistic market values, then the need for greater though not absolute accuracy has become much more of an imperative. Future value models are able to achieve greater accuracy in a more stable environment but less so when shock events occur. While the Master Appraisers will still not be able to establish the timing and extent of shock events, past events increasingly allows a greater understanding of such events to establish a more realistic outcome, even if that then creates a variation to previous forecasts. The impact of low inflation in recent years is only just being incorporated into many projections as is the virtual collapse of large business jet values in the last two years. For those that rely on future value forecasts, the one most obvious issue is that in many cases, residual value projections have been overly optimistic – that the actual outcome has been less, sometimes significantly less, than the forecast. This is partly due to limited historical evidence but as more data and the impact of various events isolated and incorporated into organic models, then accuracy will slowly improve.