American International Group was the first company bailed out by the US financial authorities in September 2008. This action introduced a bias into the link between AIG share price and defining CPIs, which existed before September 2008. The model listed in Table 1 (see original paper) is likely to be inappropriate as related to the link and not a robust one. The defining CPIs for the December 2009 model are as follows: the index for food away from home (SEFV) leading by 1 month and the index of prescribed drugs (PDRUG) leading by 13 months. In another post, we investigated the evolution of the bet-fit model for AIG from May 2008 to December 2009.

So, the best-fit 2-C model for AIG(t) is as follows:

AIG(t)= -187.7 SEFV(t-1) + 40.9PDRUG(t-13) + 673(t-2000) + 19957

The predicted curve should lead the observed price by 1 month with the residual error of $96. In other words, the price of an AIG share is completely defined by the behaviour of the two CPI components. Figure 1 depicts the observed and predicted prices, the latter shifted one month back for synchronization.

The model does predict the share price. Therefore, it should be revisited at a monthly basis. In February 2010, the price had to be at $270. In March 2010, the price is expected at $452. On 26th of March, the price was at $34. Apparently, the model is highly biased by the bailout and can hardly give an accurate prediction. One should wait before the price will really go up to get an appropriate model.

So, the best-fit 2-C model for AIG(t) is as follows:

AIG(t)= -187.7 SEFV(t-1) + 40.9PDRUG(t-13) + 673(t-2000) + 19957

The predicted curve should lead the observed price by 1 month with the residual error of $96. In other words, the price of an AIG share is completely defined by the behaviour of the two CPI components. Figure 1 depicts the observed and predicted prices, the latter shifted one month back for synchronization.

The model does predict the share price. Therefore, it should be revisited at a monthly basis. In February 2010, the price had to be at $270. In March 2010, the price is expected at $452. On 26th of March, the price was at $34. Apparently, the model is highly biased by the bailout and can hardly give an accurate prediction. One should wait before the price will really go up to get an appropriate model.

Figure 1. Observed and predicted AIG share prices.

**References**

Kitov, I. (2010). Deterministic mechanics of pricing. Saarbrucken, Germany, LAP Lambert Academic Publishing.

**Disclosure:**no positions