Throughput Accounting: Capital Budgeting (#45)
/In this podcast episode, we cover the flaws in a traditional capital budgeting system, and how to improve the situation by incorporating constraint analysis concepts. Key points made are noted below.
The Traditional Capital Review Process
Now, in a traditional capital budgeting environment, everyone throughout the company looks around at their areas of responsibility, and decides if they need to replace equipment, or add to what they have, or use more automation. Then they write down each request on a capital request form, including an analysis of whether the request will result in enough positive cash flow to exceed the company’s cost of capital.
This form then goes to a review committee, which allocates funds based on which projects will generate the most cash, and they do that until all funding is used up.
Once you understand how throughput accounting works, you’ll realize that this approach is all screwed up, for several reasons.
First, there’s no consideration of how each project request fits into the entire system of production. Instead, each request is designed to optimize a local work center that may do nothing to increase the total throughput of the company.
Second, no one has any clue where the bottleneck is even located, so the decision makers who’re allocating capital have no idea where they should invest funds.
And third, using projected cash flows is an incredibly inefficient way to allocate funding. The reason is that these projections are really hard to derive, and reality may be so far off the original projection that it’s downright funny. And for that matter, some managers make up their cash flow projections out of thin air, just to get their projects approved.
In short, if you use a traditional capital budgeting system, it’s a minor miracle if you end up investing money in the right places, so that throughput actually increases. Most of the time, all you’ve done is invest money to improve the efficiency of areas that have nothing to do with the bottleneck, so that your return on investment actually declines.
Capital Budgeting Using Throughput Analysis
So, let’s do capital budgeting the right way. Your first goal is to invest only in those areas that improve the capacity of the bottleneck. You do this by comparing the incremental additional throughput created by making an investment to the amount of the investment and any related change in operating expenses.
You should also spend a lot of time reviewing any requested improvements to the bottleneck operation. Just because you now want to invest money in it, that absolutely does not mean that you install a massive, fully automated machine there. Automated equipment has a bad habit of breaking down, which completely wipes out your throughput. Instead, it may make a great deal of sense to invest in multiple, lower-end machines that require a lot less maintenance. That way, the bottleneck may now be comprised of - let’s say - ten concurrently operating machines – so if one or two of them needs to be down for maintenance, no big deal. Most of the machines are still operational, and you still have throughput.
Just to pound on that topic a bit more, it is generally a bad idea to replace less complex equipment with more complex equipment at the bottleneck. Whenever you do that, the risk of equipment failure increases, and this is the wrong place to have more risk.
As you may recall from the last episode, it also makes sense to selectively increase capacity in some upstream areas, because this gives you more sprint capacity. If you have areas like that, then absolutely consider investing in more capacity. Especially when you’ve had cases where they didn’t have enough capacity to catch up from a system crash, and it impacted the bottleneck.
However, this does not mean that you need an overwhelming amount of sprint capacity. Just analyze the situation, and base the investment on the capacity levels at which the equipment is currently running.
Also, if someone puts in a request for funding in an area that has no impact on the bottleneck, then basically beat the hell out of it – especially if it’s for a lot of money. That means interviewing them about why they need it, requiring extra levels of approval, examining the impact on the entire system, and so on. Now, this doesn’t mean that you’ll automatically shoot down every non-bottleneck request. There are lots of cases where a non-bottleneck investment can mitigate risk, or it may reduce operating expenses. But, you should absolutely dig into these proposals to make sure that they’re really necessary.
By viewing capital budgeting from the perspective of throughput analysis, you get some really interesting results. Obviously, overall throughput will probably increase. What’s less obvious is that total investment may drop by an incredible amount.
Now, you may think – sure it will drop for a year or two, but what about in a couple of years, when all of those non-bottleneck operations start to wear out, and need to be replaced? Well, when that time comes, you view replacement investments in terms of how much capacity you really need, versus how much you had. Chances are, you can invest in less capacity and less automation, so that when a large machine wears out, you can replace it with one or more smaller machines that are less expensive and easier to maintain.
And a few additional thoughts on capital budgeting. Under the traditional system, this is a once-a-year event. But realistically, if the throughput level is suffering because of a missing investment, just do whatever it takes to make the right investment right now. It’s completely ridiculous to only make an investment decision once a year, because you may lose a bunch of throughput dollars by waiting.
Also, you normally hand off a capital request to a financial analyst, who verifies the reasoning behind the cash flows in the request. But under a throughput-based approach, your main concern is how the capital request will impact the bottleneck, and that calls for a review by a process analyst, like an industrial engineer. Isn’t that interesting?
And finally, you need to change the capital request form. Instead of having a big space in the middle for a cash flow analysis, ditch that part entirely. Instead, divide it into three boxes. You fill out the first box if an investment will improve a bottleneck operation. And in this box, you enter the impact of the investment on throughput, operating expenses, and return on investment. This box requires the smallest number of approvals.
You fill out the second box if a project will mitigate risk. If so, itemize exactly what risk you’re addressing. This box has a different set of approvals, including the chief risk officer, if you have one.
And finally, there’s the third box, which you fill out for non-bottleneck investments. This one requires the most comprehensive justification, and – as you might expect – it requires the most approvals.
So, there you have it. A completely different way to handle capital expenditures, and it’s all driven by throughput accounting. In the next episode, I’ll cover how to conduct financial analysis using throughput concepts.