SalesForekast™ uses historical sales (or market cap) data to determine which macro-economic indicators correlate, and models that into the future. In this post, the details behind the model will show how it works.
We uploaded the following data file: Demo Financial Data
The variables with highest correlations, up to six, are then narrowed down by an iterative process that ensure each variable is at least 95% significant (until all remaining variables have a p value below 0.05). If no variable has correlation of at least 85%, the software will report ‘no correlation’. (Note: the .85 number can be manipulated to meet the user’s specification)
The output of SalesForekast™ provides the regressions statistics for each successful regression.Once a regression is performed, the regression statistics are available as below:
In this case, the multiple R value was 90.1%. This is the probability of the actual sales lying within range indicated. The standard error was 0.345 - or about 4% of the quarterly sales.
However, to be clear, the goal of SalesForekast is not to absolutely predict the future. The software’s primary benefit is to provide an alternative view of the future based on correlations with macro-economic indicators. Most companies do not look at this very critical variable.
If trends in GDP, the consumer price index, employment levels, construction spending or one of 25 other such variables impact sales, it is highly important to consider this in future projections.
Traditional bottoms up forecasts do not consider this variable. Rather, they focus only on sales trends and may adjust for new products and new customers.
Changes in the macro-economic environment do have an impact, as we have seen several times in the past. For example, large global companies have panels of dozens of economists to provide senior management insight for this same data.
What does SalesForekast™ do for your company?
It identifies which macroeconomic variables, if any, impact historical sales (or market cap or P/E for public companies). This is great insight on what is important for your business (or division, geography, customer, etc).
It enables you to test sensitivities on macroeconomic variables. If consumer price index has a significant correlation, different levels of projected growth to see how big the impact might be ($X change in sales per one point change in the CPI).
If none of the 30 variables in the mode have a statistically significant impact, you have access to the Federal Reserve Economic Database of over 500,000 data tables to see if any other variables correlate.
The forecasts developed by this software should primarily be used to adjust bottoms up internal forecasts.
The software allows you to tests forecasts for companies, divisions, or product lines. The more cohesive the business, the more likely that significant correlations exist.
It can be used to screen potential acquisitions as most major private equity firms and large strategic buyers use this methodology late in the process because it has traditionally been very expensive.
Last, SalesForekast™ can help determine whether investment in a public company is warranted. This software gives you real insight that others will not have - and you can use your personal knowledge of likely future changes in say GDP or M1 funds to build your own portfolio (e.g. invest only in companies that are correlated to infrastructure spending or to household formation - both expected to increase).