All project managers or investment analyst have been there: We prepare a financial model as part of a business case to justify a project or a new business venture and we all have to ask Finance for the hurdle rate or the discount rate, depending on which technique we are using: IRR or DCF. And they give us a number for us to use it until they tell us otherwise. Or they tell us that they have given it to us before. That's how it works in most companies, there is only one hurdle rate or discount rate that is used by everyone until it is changed by Finance. But does that make sense?
The textbook way of working out hurdle/discount rate (let's just call it the discount rate) is to base it on the weighted average cost of capital. So it changes when there are changes in the company's share price (affecting the cost of equity), the base lending rate (effecting cost of debt) or a new large loan is taken out or shares issued (affecting the weightage). This ignores a couple of factors.
First, there is the investment horizon. Giving us a single discount rate implies that the cost of capital is stable over the entire period the investment is intended to provide a return. We know that the cost of debt is based on the yield curve over time: longer dated debt cost more than shorter dated debt, all other things being equal. So, should we have a discount rate that works like the yield curve? Also, cost of funds for equity in short term (say as demanded by a speculator) could be much higher than that in the long term (say as demanded by a value investor). While scholarship in this area is murkier than for the cost of debt, I would think the difference would be significant. A philosophical question: is the investment decision on a project a proxy for the investment decisions of the ultimate financier, which is what using the weighted cost of capital seem to imply; and does the answer have any impact on the discount rate we use.
Another point: is the discount rate used intended to match the weighted average cost of capital in each year of the financial model because future cost of funds could change every year? If so, would we need to adjust the discount rate for each year, depending on the expected cost of debt & equity and the projected weightage in each year? Meaning, Finance will have to give us 31 discount rates for a 30 year financial model, including one for Year 0 (I mean if you want to be pedantic: there's a risk premium in month 11 that doesn't exist for month 1)?
And yet another point: the cost of capital worked out covers only the financing risk. There is also the projection risks. Numbers used in Year 30 of the financial model would of course be much more uncertain than in Year 1. Should that increased uncertainty be reflected in a higher discount rate for Year 30 than for Year 1 even if the weighted average cost of capital for some reason remains the same? After all, there is an extra risk premium for Year 30 compared to Year 1and I always find that reducing numbers downwards is too blunt a tool to handle uncertainty: prudence is a concept for accountants as they are OK to err in only one direction but investment analysts need to be accurate and loses out if the erring is in either direction, depending on whether you are buyer or seller.
Another factor would be the risk of the project. A project to replace the HR software and another for a factory in a new market cannot have the same risk profile. So, how many risk gradations would there be? Projects of an operational nature would clearly have different risk profiles from business expansions. But even so, risk profiles defies standardisation. Installing a CRM system where none existed would be much more complicated with uncertain benefits (and even costs) than upgrading the accounting software.
And expansions into new markets depends largely on how new is new. A new factory to expand the capacity of the existing phone factory next to it is vastly different from a new phone factory for a formerly rubber business (ala Nokia). Moving into a overseas market again has a different risk profile but again depending on (a) whether the company itself had prior experience in overseas investment anywhere else, (b) whether the set up requires a foreign partner and how reliable is that partner, (c) forex and many other risks associated with foreign investments and managing a foreign project, and in which country too. It all depends on how far from the legacy business it gets.
And shall I get personal here? We have all known that one manager who cannot be trusted to run a project - so should we adjust the discount rate on account on increased risk? Or that business owner who is always wildly optimistic in their projections? (I was always in favour of comparing business owners/departments actual benefits/cost against their submitted business case in order to build up the history of their over/under optimism so that it can be adjusted for when they next turn up for approval.) OK, I am probably going too far here but you get the gist: the discount rate should include all risk premiums and the discount rate should differ if the risk profile of the project differ. It is after all based on the cost of funds, and the question is how much interest would the bank themselves charge if they were to be funding that project directly by itself, not just lending to the company.
So, what do we do? Does this mean we need to swing from having a single static companywide discount rate to the other extreme where we submit all the details to Finance and wait for them to give us the discount rate? I tell you what would happen: first, the analyst would go ahead with a best-guess discount rate anyway for initial evaluation before gathering enough details to feed into the project initiation process, followed by almighty arguments over Finance's evaluation of the risk profile of the investment that would make your dispute over budget allocation seem (umm) friendly. Let's not forget that accountants think they are paid to be risk-adverse.
So, maybe there should be several scenarios for which discount rates could be pre-assigned: (say) replacing existing software, all the way to expansion into totally new markets. The lowest discount factor would be for maintenance operational projects, pegged at just below the corporate cost of fund. This could be adjusted for longer investment horizons. Finance may be allowed to adjust the final discount rate up or down within certain limits on risk assessments (I am a bit sceptical here though, as experienced project managers will then make the financial model work using the higher discount rate, pushing projections to further reaches of the envelope to do so). Anyway disputes over the adjustment will be adjudicated by the CEO & CFO but let's face it: such adjudications that would have made a difference to the approval outcome would be as rare as adjudications over US presidential election ballots that would make a difference to the election outcome.
But you know what about the above proposal: pardon my cynicism but too many Finance departments are just too much bean-counters to have the agility to pull it off. Maybe a few. Maybe. But if you are one of those, would like to hear from you. There could be some out there.
Comments