Articles

Should a company build, buy or partner for a new capability?

Three routes to the same capability cost different amounts and fail in different ways. Weight each by its chance of working before comparing the present values.

A company wants a capability that would add $30m a year of cash flow. Built in-house it costs $40m, takes three years and works 60% of the time; bought it costs $120m plus $15m of integration; licensed from a partner it costs $12m up front and half the cash flow for six years. At a 10% discount rate the risk-weighted values are $32.3m to build, $23.1m to buy and $38.5m to partner. Without the weighting, building looks best at $76.2m.

The assumptions

Illustrative capability, invented figures, $m
ItemBuildBuyPartner
Up-front costnone12012
Spend after year 015, 15 and 10 in years 1 to 315 of integration in year 1none
Cash flow to the company$30m a year, years 4 to 10$30m a year, years 1 to 10$15m a year, years 2 to 7
Chance it works60%85%85%
Discount rate10%10%10%

The partner takes half the cash flow and the licence ends after year 7, when the company would have to build or buy a replacement. A failed build still costs all of its spend; a failed purchase still costs its price. That asymmetry is why the chance of success belongs in the comparison.

Step 1: the present value of each cash flow stream

Build: spend = 15 ÷ 1.10 + 15 ÷ 1.102 + 10 ÷ 1.103 = $33.5m; cash flow = 30 for years 4 to 10 = $109.7m

Buy: cash flow = 30 for years 1 to 10 = $184.3m; integration = 15 ÷ 1.10 = $13.6m

Partner: cash flow = 15 for years 2 to 7 = $59.4m

Excel: =NPV(10%,0,0,0,30,30,30,30,30,30,30) for the build cash flow

Step 2: weight by the chance of success

Value = chance × PV of cash flow − certain costs

Build = 0.60 × 109.7 − 33.5 = $32.3m

Buy = 0.85 × 184.3 − 120.0 − 13.6 = $23.1m

Partner = 0.85 × 59.4 − 12.0 = $38.5m

Value of each route, $m
RouteIf it worksRisk-weighted
Build76.232.3
Buy50.723.1
Partner47.438.5

If every route worked, building would lead at $76.2m, ahead of buying at $50.7m and partnering at $47.4m. Once the chances are applied the order is reversed: partnering leads by $6.2m over building and $15.4m over buying, because it puts almost no money at risk before the capability exists.

Step 3: the break-evens that change the answer

Build overtakes partner when chance × 109.7 − 33.5 = 38.5

Chance = (38.5 + 33.5) ÷ 109.7 = 65.6%

Buy overtakes partner when 0.85 × 184.3 − 13.6 − price = 38.5

Price = $104.6m

What if: a surer build, a cheaper target

Risk-weighted value when one input changes, $m
CaseBuildBuyPartner
Base case32.323.138.5
Build works 75% of the time48.823.138.5
Target bought for $100m32.343.138.5
Licence runs ten years (years 2 to 10)32.323.154.8

Each change reorders the routes. At 75% the build wins, at $100m the purchase wins, and a longer licence widens the partner's lead to $22.5m over the build.

The common mistakes

Takeaway

Questions readers ask

How do you compare building, buying and partnering?

Discount each route's cash flows, multiply the cash flow by its chance of working, and subtract the costs that are paid regardless. Here that gives $32.3m, $23.1m and $38.5m.

When does building beat partnering?

In this example when the in-house project works more than 65.6% of the time. At 60% the build is worth $6.2m less.

Why does the purchase look worse after weighting?

Its price and integration cost, $133.6m in present value, are paid whether or not the capability delivers. A 15% failure rate removes $27.7m of value.

The workbooks behind this article

The free companion files for The Corporate Development Professional rebuild a fictional acquisition's standalone value, synergies, walk-away price, accretion and return on capital with live formulas. The capability in this article is invented and is not in them. No account and no email address.

Open the companion files →

Not on Amazon: 36 paperbacks by Julian R. Sterling on how to read the documents, printed on demand and shipped worldwide. Browse the bookshop →

Also on this site

Reading guide: corporate finance, valuation and markets → · All 1908 articles →