"Leasing saves cash" and "buying is cheaper over time" are both true and both useless on their own. What works is putting both sides into one formula and then seeing which variables the conclusion is sensitive to.

1. Set the formulas

Three-year cost of ownership = purchase price − residual value after three years + maintenance + downtime loss + cost of capital + amortised site and power works − depreciation tax benefit (per applicable policy)

Three-year lease cost = cost of capital tied up in deposit + monthly rent × 36 + over-usage charges + amortised works borne by lessee + restoration cost at exit

Put both on the same timeline and express both as three-year cash outflow plus end-state asset position. Otherwise you end up comparing incompatible bases.

2. Five costs that must be counted

CostBuyLease
Cost of capitalFull amount, at internal rate or loan rateDeposit only
Residual value riskBorne by owner; used-market swings hit the conclusionBorne by lessor
Technology changeOwner carries it; new model means new capexModel swap possible after term, if negotiated
Maintenance and sparesIn-house capability or separate service contractUsually inside rent; confirm response time and scope
Downtime lossUnchangedUnchanged — rent does not cover lost output

3. Three decisive tests

Utilisation. The more effective running hours per year, the more fixed cost buying absorbs. Low-intensity, intermittent use usually favours leasing.

Pace of technology change. If there is a real chance the process itself is replaced within three years — manual feeding becoming automated feeding, for example — leasing transfers obsolescence risk, and that is often worth more than the rent premium.

Cash flow and financing cost. Buying is a lump sum; leasing is spread. Where capital is expensive or credit lines must stay available, leasing has the lower opportunity cost. Where cash is strong and depreciation rules are favourable, buying wins.

4. Run sensitivity, not a single point

Turn three variables into ranges: annual running hours, residual rate after three years, probability of process change. Then find where the conclusion flips. A conclusion that holds across the plausible range can be acted on. One that holds only in a narrow assumption set is a high-risk decision, and a hybrid structure deserves consideration.

5. Hybrid is usually best

Buy the core process — long-term, stable, mature. Lease peak capacity, pilot validation and temporary projects. Fixed cost then sits only on confirmed demand, variable demand is bought with rent, and the option to change models is preserved.

Three clauses are most often overlooked in a lease: how early-termination penalties are calculated, who owns modifications (added fixtures, changed programs) and what restoration is owed, and whether rent includes software licences and remote support.