I have no idea if I'll keep going with this, but all things start somewhere. I am working in the field of corporate decision-making, decision analysis and finance. My current focus area is portfolio analysis.
The traditional approach to corporate decision finance is to forecast a series of numbers for revenue, costs of sales, research and development and capital spending and then discount those cash flows to come up with an NPV for the project. Those projects are put in NPV order and a line is drawn when the business runs out of budget.
This approach leaves a lot to be desired. For one thing, forecasting is hard. Point estimates of just about anything in the future are likely to be wrong. We can solve some of that problem by using range estimates and Monte Carlo analysis. We still need good range estimates, however, and that means relying on either historical regression or experts (or both). Each of these have their own short-comings, and techniques for compensating for them.
Another problem with this traditional approach is the shortcoming of the NPV approach itself. A simple thought experiment helps to highlight this:
Imagine a collection of projects:
Project 1: NPV 100, spending 30
Project 2: NPV 70, spending 15
Project 3: NPV 55, spending 10
Project 4: NPV 50, spending 10
Project 5: NPV 45, spending 9
Project 6: NPV 40, spending 8
Project 7: NPV 35, spending 7
If we have a budget of 45, the traditional approach would recommend the top two projects (the highest two NPV projects) for a total NPV of 170.
There are other recommendations that make better sense. If we instead recommended the bottom five projects, we would also spend 45, but would have a total NPV of 175.
The real world is far more complex than even this. Unlike a portfolio of financial assets, most business projects have complex interrelationships with one another. Most businesses have synergies between their different units and add incremental projects to their portfolio of activities because they may have indirect benefits to other projects. For example, a company that offers a complete suite of personal efficiency software will find that each individual product sells better as part of the whole than any part would if they only offered one or two products (network effects, bundling). Another example is the idea of cannibalization.
Big businesses often have organizations dedicated to R&D. These groups develop new technology and create the building blocks that other divisions might bring to market. Often these building blocks have complicated relationships with one another and with the products that they go into. Valuing these sorts of efforts from a financial decision-making approach has traditionally been accomplished by doing some sort of allocation or amortization of the spending to the child products. This approach is inferior, and leads to wrong answers.
I have been working on a new approach to corporate financial decision-making. In the upcoming weeks I will explore this approach on this blog (or at least try to as time permits). I plan on skipping subjects that are extensively discussed elsewhere, such as the basics of finance, economics and strategy. I will write about the fundamental elements of financial analysis, a bit about real options and incremental value and some about biases and heuristics. I'll write a lot about portfolio and portfolio valuation and decision-making.
The traditional approach to corporate decision finance is to forecast a series of numbers for revenue, costs of sales, research and development and capital spending and then discount those cash flows to come up with an NPV for the project. Those projects are put in NPV order and a line is drawn when the business runs out of budget.
This approach leaves a lot to be desired. For one thing, forecasting is hard. Point estimates of just about anything in the future are likely to be wrong. We can solve some of that problem by using range estimates and Monte Carlo analysis. We still need good range estimates, however, and that means relying on either historical regression or experts (or both). Each of these have their own short-comings, and techniques for compensating for them.
Another problem with this traditional approach is the shortcoming of the NPV approach itself. A simple thought experiment helps to highlight this:
Imagine a collection of projects:
Project 1: NPV 100, spending 30
Project 2: NPV 70, spending 15
Project 3: NPV 55, spending 10
Project 4: NPV 50, spending 10
Project 5: NPV 45, spending 9
Project 6: NPV 40, spending 8
Project 7: NPV 35, spending 7
If we have a budget of 45, the traditional approach would recommend the top two projects (the highest two NPV projects) for a total NPV of 170.
There are other recommendations that make better sense. If we instead recommended the bottom five projects, we would also spend 45, but would have a total NPV of 175.
The real world is far more complex than even this. Unlike a portfolio of financial assets, most business projects have complex interrelationships with one another. Most businesses have synergies between their different units and add incremental projects to their portfolio of activities because they may have indirect benefits to other projects. For example, a company that offers a complete suite of personal efficiency software will find that each individual product sells better as part of the whole than any part would if they only offered one or two products (network effects, bundling). Another example is the idea of cannibalization.
Big businesses often have organizations dedicated to R&D. These groups develop new technology and create the building blocks that other divisions might bring to market. Often these building blocks have complicated relationships with one another and with the products that they go into. Valuing these sorts of efforts from a financial decision-making approach has traditionally been accomplished by doing some sort of allocation or amortization of the spending to the child products. This approach is inferior, and leads to wrong answers.
I have been working on a new approach to corporate financial decision-making. In the upcoming weeks I will explore this approach on this blog (or at least try to as time permits). I plan on skipping subjects that are extensively discussed elsewhere, such as the basics of finance, economics and strategy. I will write about the fundamental elements of financial analysis, a bit about real options and incremental value and some about biases and heuristics. I'll write a lot about portfolio and portfolio valuation and decision-making.
