# Facing into M&A *Price growth before you shop for it* --- Acquisitions get a bad rep. Value-destroying, dangerous, distracting for management, bad for shareholders: the charge sheet is long, and the research behind it is real. I've bought a few, so I've read that research with some interest, and the striking thing is what it actually studied. Nearly all of it is one kind of deal - Fortune 500 buyers, transformative mergers that take years to digest and sometimes never do. The finding is sound. The advice drawn from it travels a great deal further than the evidence does, and almost none of it is about a £100m business buying a £10m one. Below a certain size the sceptics are simply right. A business without the scale and maturity to absorb another one's complexity will find a deal massively distracting and overwhelming, whatever the strategic logic said. Organic growth really is the better route for most companies most of the time. Somewhere north of £50m of revenue, the arithmetic starts to change. Revenue is a proxy for two harder questions, and it's the questions that actually bind: could you lose a senior person to an integration for a year without the core suffering, and could you write the whole cheque off without breaching anything. Around that size, the core lines are at a reasonable level of maturity, and maturity is a polite word for the fact that they won't carry the next leg on their own. Getting to the next level means going somewhere you aren't already: new channels, new markets, new geographies - provided the destination is somewhere you could actually come to own, rather than open water. You can build your way in, which is slower and often right. You can stop growing, run the core for cash and hand it back to whoever owns you, which is a perfectly respectable answer that most writing on this pretends doesn't exist. Or, in a lot of instances, it's a genuine build-versus-buy decision, and a sensibly sized bet gets you into a new area faster than building. ## Which deal is this? Every small acquisition is one of two deals, and most of what goes wrong comes from pricing one and running the other. A leave-alone deal is roughly what the case says it is. You inherit the customers, the operations, the team and the working pattern, you collect the cash flow, and the integration work is paperwork: bank accounts, group reporting, audit harmonisation, a few weeks of finance time and you're done. The upside is bounded by what the business already does, and so is the risk. A change-the-business deal is different, and it's where the strategic logic usually lives at mid-market scale. The reason the bolt-on in the adjacent vertical is interesting is rarely the cash flow at the asking price. It's that you can put your sales team into a market they couldn't reach on their own, or run the acquired customers through your operating processes, or put your channel behind a product you already understand. The upside is real and bigger than the leave-alone case, and so is the work: a change-the-business deal is a turnaround you have gone out and bought. The part I underrated, the one time I bought one without meaning to, was the people. We felt hamstrung by a few key individuals we didn't think we could manage, or overmanage, for fear of them walking out and leaving us with even more to do. Both can work. Being honest about which one this is matters because the price and the terms are very different. If the leave-alone version is worth £25m, the change version is rarely worth more, because the extra upside is created by your work after close, and you shouldn't pay the seller for value you're going to have to build yourself. The exceptions are rarer than deal papers suggest, and [[Cintas|route density]] is the honest one. ## What growth is worth Either way you are buying growth, so it's worth knowing what growth is worth before you pay for it. Boards mostly take the cost programme over the expansion, and I understand why. The two papers sit in the same pack: the chief operating officer wants a programme that takes about £1.3m a year out of the base and drops it straight to profit, banked in the run rate, from a team that has done it before; the commercial director wants growth up from 5% to 8% in a market the business has never sold into, and the return on it is the number nobody in the room can evidence yet. Say the business makes £10m after tax, earns 16% on its capital, and its investors would settle for 10%. On a perpetuity it's worth £138m as it stands. The cost programme lifts profit to £11.3m and the return on that capital to 18%, taking it to £162m, so it creates £25m. Growth at 8% takes it to £250m: £112m created on the same £10m of profit with no cost saving at all, and about four and a half times as much. It isn't free. Growing at 8% rather than 5% means reinvesting half the profit rather than a third, so the cash you can take out drops by roughly £2m a year, and the £112m is what you're being paid for giving that up. <iframe src="https://anishshailpatel.github.io/instruments/margin-or-growth.html" title="What a cost programme and an expansion are each worth" width="100%" height="1560" style="width:100%;border:0;display:block;margin:1.2em 0;" loading="lazy" sandbox="allow-scripts allow-same-origin"></iframe> Growth wins through the spread. Every pound reinvested earns six points more than investors require, and growth puts more pounds to work on that spread each year, on a larger and larger base, where a point of margin arrives once and lands in the run rate. Take the spread away by setting the new market's return to the 10% hurdle and the curve goes flat, because every extra pound is then earning precisely what it costs. Below the hurdle the arithmetic runs in reverse and growth destroys value while looking, in the management accounts, like success. Which is why the case rests on three numbers and the paper usually shows one. There is the growth rate itself, where the denominator is only two points wide, so 7% instead of 8% turns £112m of value created into £50m, and at 6% the cost programme was the better call all along. There is the return the new capital earns, which is what actually decides it: at 12% against a core earning 16% the expansion is worth £167m against the cost programme's £162m, and below about 12% the cost programme wins outright. And there is how long the faster growth lasts, which the perpetuity hides completely: 8% for five years and then a fade back to 5% is worth £144m, less than the cost programme, and it takes about sixteen years at 8% to draw level. That last one is the number I'd want a paper to defend hardest, and it's the one nobody puts in. The model is being kind to the cost programme too, mind: it lets the whole £1.3m arrive in year one and stay out for good, where in practice savings have a habit of leaking back. > [!note]- The working > $\text{value} \approx \frac{\text{free cash flow}}{\text{required return} - \text{growth}}$ > Free cash flow is profit after tax less the reinvestment growth demands, and the reinvestment rate is growth ÷ return on capital; on its own that is a perpetuity. For a finite run the instrument lets the faster growth and the new-market return hold for the years you set, then drops growth back to today's rate at the core's return and discounts the rest back. It shows the three cases side by side before it draws the curve. So what you are paying a premium for is access to returns above your hurdle, in a market you couldn't reach organically at a sensible cost, held long enough to matter. That is a lot to ask of a business you have just met, which is why the price you pay for it wants building from the bottom. ## Price the deal you'll actually run A business that's been prepared for sale has had its costs trimmed, its working capital pulled in and its capex deferred, all perfectly legal and routine. Where the founder has been paying themselves below market for a couple of years to dress the business, the profit you're buying is smaller than the profit in the information memorandum. Against a broker's glorified estimates, the number can very easily move twenty-five to fifty per cent once more realistic assumptions are embedded, which is usually enough to change the answer rather than merely adjust it. The board pack will typically handle all this with a single risk haircut on the headline price. A percentage off the top is a gesture at uncertainty; what you want is the case rebuilt on normalised earnings, with the founder's real cost of employment in it, the working capital where it will actually sit, and the deferred capex back on the line. A haircut lets everyone keep the number they arrived with. I've owned a business whose profit turned out to be held up by two contracts priced far above what the work was worth, quietly subsidising a lot of other work that was marginal at best. We had discussed it with the vendors and taken warranties on it. Rebuilding the case on normalised earnings would have found it. The haircut we were arguing over would not have, and we spent the next two years discovering the difference. Then price in what the deal does to you. Senior management spent on integration is senior management not spent on the core, and if the model assumes the core grows at trend through the transition, the model is wrong. Build the version where the core flatlines for eighteen months and see whether the deal still clears. Write the post-close plan before close, with names and dates: *integrate sales by Q3, harmonise the operating processes by Q4, hold pricing through the transition* is a plan; *explore opportunities to leverage the platform* is not. The plan will be wrong by week six. Writing it down is how you notice quickly which parts are wrong. That is most of what a plan is for. And structure the terms to hold the seller to the business you intend to own rather than the one they've spent two years preparing to sell: deferred consideration, earn-outs tied to retained customers, rollover where the relationship matters. Warranties help less than you'd hope. I've won a claim under them: it took two years, we got some of the money back, and we were still stuck with a business that wasn't what we thought it was. ## Size it so being wrong is survivable Then size it, and not for the usual reason. The bets have to be much smaller than the core: small enough that a bad one gets absorbed and the business carries on. The standard defence is portfolio logic: do a lot of them and some failing is part of the plan. That works for a serial acquirer like Halma, which has done small deals for decades, keeps every one below a threshold, and turns down anything that would force a new operating model on it. It doesn't work for you. A mid-sized business does two or three deals in a decade, which isn't a portfolio, and the averaging that makes the portfolio argument work never happens. Each deal has to clear on its own merits. Small enough to survive stops being a portfolio argument and becomes a condition on every deal you do. > [!note]- The working > The underlying point is about [single runs versus averages](https://www.nature.com/articles/s41567-019-0732-0). A bet that's positive on expected value across a thousand parallel worlds can still ruin the one player who takes it repeatedly in sequence, because a big enough loss removes your ability to keep playing. When you only get one run, the size of the downside matters separately from whether the odds are good. Turning down positive-expected-value bets that could end the game is the correct arithmetic for someone with one timeline. ## What the capability is for If those conditions hold, deal literacy is part of the operating job rather than something the corporate-finance people do. Being able to look at a business and form a view on what it's worth, and being comfortable enough with a process to know when it's going badly, is a genuinely attractive thing to be good at. Be clear what the capability is for, though: pricing properly, and saying no quickly. Most of the times you use it, that's what it will be doing. ---