Taxes and Depreciation
Prev Next

Taxes and Depreciation

Depreciation is the allocation for tax purposes of the purchase costs of fixed assets.

IRS allows depreciation of investments which means each year a portion (based on a depreciation schedule) of the original costs is considered an expense reducing taxes due. Most Data Analytic projects are more profitable at the beginning of their life and with few exceptions the profitability degrades quickly. Competitors will learn fast how to copy the source of success, copycat projects would use cheaper and more powerful hardware meanwhile the older hardware will be more expensive to maintain. Besides IRS allows for depreciation only of the original cost ignoring inflation. The portion of cost depreciated later brings a less valuable contribution to the cash flow. Therefore it makes sense to depreciate the investment as fast as possible. In the field of Data Analytics the most common depreciation schedule is the Modified Accelerated Cost Recovery System (MACRS) structured over three years (33% first year, 44% second year, 15% third year, and a residue of 8% the forth year). Less common but not unheard of are the five and seven year depreciation schedules.

If the project brings revenues over costs the investment is recovered from taxes through depreciation. It increases the profitability but has the perverse effect of increasing costs. To understand this issue imagine you have to make a decision: after spending one million dollars your team has discovered a major issue in the current architecture. You can continue the project with the internal resources at a cost of $100,000 and 25% chance of failure or acquire a library at $250,000 and negligible chance of failure. Most companies would choose to buy the library which in most cases makes economic sense but that would add to the costs.

Taxes are applied to Revenue minus Cost and Depreciation multiplied by the tax rate (aka EBIT - Earnings Before Interest & Taxes).