Projections
I am inherently distrustful of projections. For too many years as an investment banker I reviewed business plans and their projections. If you asked any of the authors of the projections- no matter what the descriptors use- one descriptor was always used “conservative”. All the projections were conservative. No projections were ever described as realistic, fantastic, or even outlandish. They also had a feature of always starting in the lower left and rising to the right at a consistent slope when being charted. From my memory the scale of the graph also seemed to be manipulated so that the line was as close to possible, a 45-degree angle.
Roll the clock forward 25 years to last month. A large accounting firm was retained to craft projections for the revenue of a city. I had a flash-back moment – the unbending line from left to right at a 45-degree angle was back. A city had invested a great deal of money to get projections of future tax revenue to be able to plan the sources and uses of future revenue. Managerially a sound idea until you had a closer look at the projections. Month by month these projections inched ever upward.
So what were the problems?
The city has a well-known seasonality to their revenue. 8 months of the year the local university is in session and for 4 months the university is not in session. Month to month variance can be as large as 40% up and down! Several new buildings were being built that had significant retail space – no account was made for over 500,000 square feet of retail space coming on line had been made. No account had been made for business cycles. The economy runs hot and cold, some years it is hot and some years it is not hot at all. No “adjustment mechanisms” had been incorporated to incorporate leading indicators of either downturns or upturns in the economy – such as business permits, new, cancelled and current.
I pointed this out and suggested that the city and the accounting firm get some skin in the game. I suggested that for each year the city exceeds projections the city pays for the accounting firm the overage. Conversely, the accounting firm was paying the city for each year the city did not meet projections. There was a great deal of nervous laughter from the public gallery and a load of squirming from the city officials and the accounting firm.
The city manager then broke the silence and said, “The projects are also needed as we are going to seek tax revenue collateralized debt. Debt collateralized by future tax revenue and projections are required to be granted financing.” (Ah, not just a planning tool).
“So”, I responded scratching my head, “you don’t need an accurate appraisal – you only need an appraisal – check the boxes so to speak? Correct? Like an appraisal for a home loan – you don’t need an accurate home appraisal – just any appraisal will do?”
The public gallery roared out loud with laughter. But that was it. The city passed on the projections – accepted them as submitted and moved on. Can you imagine if you were required to turn in a report on a topic? The report does not need to be accurate, it just needs to be thick – possess a ton of footnotes, and printed on glossy paper. (Sounds like most fraudulent stock offerings I have seen.)
The validity of projection decreases in a logarithmic fashion from the date of the first inputs – month my month. Essentially projections – with static inputs have no value at all by 13 months – or sooner.
Now projections with dynamic inputs and those, which make use of leading indicators, will typically remain useful for 18 months. However, if one makes dynamic adjustments and is able to actually qualify and quantify leading indicators, one not only has a model but a useful predictive tool.
Planning for planning’s sake is useless. However, making plans that have both a feed forward and a feed back control mechanism can be very useful. Furthermore, with dynamic inputs, the projections can be revised every month and as long as you are within your acceptable margin of error (looking back about three months)- you are doing pretty good.
