Breaking the Model (Updated)
March 6, 2023 | Issue #80
Big Deal Small Business: Breaking the Model (Updated)
March 6, 2023 | Issue #80
I’m editing & updating older posts to reflect my latest thinking — this one in particular needed some updating as it referenced loan rates at 5-6%…sadly, you’re looking at more like 10-11% now.
Side note — I’d like to organize a golf outing for SMB operators & searchers in the Seattle area for sometime in May. Thinking a weekday around 11ish, followed by drinks at the clubhouse. If that would be of interest, please reply to this email!
Confirming that an SMB acquisition “pencils” is fairly easy. So building a financial model is less about figuring out the ROI math, and more about figure out what will break the model.
In this issue, I will walk through how to build & utilize an SMB model in a practical way. This will allow the model to become a crucial element of your diligence & negotiation process.
But first, why is ROI math so easy?
Because you buy them so cheap. Here’s the math in simple terms – if you buy a business for 5x unlevered cash flow (i.e. before any debt payments), that’s the equivalent of a 20% equity return. You buy $20 of cash flow for $100, you get $20/year, so 20% per year.
Then layer on 90% SBA debt at a 10% interest rate (aka $90 if you bought the business for $100). SBA loans are fully-amortizing over 10 years, so you will pay $14.65/year out of your $20 of cash flow.
You only have to put $10 of equity in, which leaves $5.35 of cash flow per year for that equity — 53.5% annual equity return before taxes.
So as long as you believe the business will stay flat year after year, the ROI math will pencil out for the equity.
BUT — the magic is in the “as long as you believe the business will stay flat year after year” concept — that’s where money is made or lost.
Creating a single model with growing or even flat earnings doesn’t really tell you anything about the risk profile of the deal. As I showed above, the math will always work out in those cases. While that kind of model can help you assess how quickly you can grow, the real threshold question should be assessing the deal’s risk profile.
Setting up the P&L
The first thing I do is input at least 3 years of P&Ls into a fresh spreadsheet by hand. Doing it manually helps you notice when something seems awry. Skimming P&Ls won’t lead you to notice that the cost of uniforms increased from $3,000/year to $5,000/year, which is a weird increase unless they hired a bunch more staff. Noticing these details helps you ask useful questions.
I sometimes even send the Seller my recast financials, with questions in the margin so they can see exactly what I’m looking at. It helps keep their answers focused and detail-oriented rather than long-winded and unhelpful.
Practically, my method is to list every single line item that they provide, but then to create some buckets to group up costs. The classic buckets I almost always use are Cost of Goods Sold, Total Payroll, Rent, Owner Expenses, Insurance, Office Expenses, etc.
The key judgement, however, is separating out variable costs from fixed costs. Given our goal is to assess the risk profile of the transaction, we need to understand what costs automatically go down when times get tough, versus which costs require real belt tightening to reduce.
Variable costs can be variable to either revenue dollars or volume, so make sure you understand the distinction. For example, if you pay your salespeople a commission based on sales, then a 3% decline in revenue = 3% decline in sales commissions.
By contrast, suppose it costs you $100 in materials to complete a job (i.e. one unit of volume) that generates $1,000 of revenue (so 10% variable cost). If the price of the job declines to $900, your material cost is actually fixed, not variable, given the price of your materials haven’t changed. It’s only a 10% variable cost if your volume is what is varying.
Anyway, once you’ve set up the detailed P&L, you can summarize it quickly into the buckets outlined above. I generally put most variable costs into COGS and most fixed costs below the gross margin line.
Setting up the Projection Model
Now that you’ve set up three years of historical financials into a summarized way, you can create a very simple model using the following rules (with the example of a tutoring business):
Break up revenue into its key drivers. For a product-based business, that could be price x unit volume. A tutoring business would be hours of tutoring provided x average $ per hour price.
Create forward drivers for price & volume, generally just % increase (or decrease) each year going forward. For the tutoring business, you would have one % line for “change in tutoring hours provided” and one for “average $ per tutoring hour”. That will allow you to generate a future revenue estimate for each year.
For variable costs, calculate what the cost is as a % of revenue or $ per unit of volume. A tutoring business would be tutor payroll costs per hour of service, or Google ads as a % of total revenue.
For fixed costs, look at recent trends to project year-over-year increases. For example, your office staff payroll likely increased 2-3%/year due to inflation historically, but may need to increase 5-8% going forward given recent inflation.
With those steps, you have a high-level model that can project earnings for the next few years.
The last element to add is your starting cash and debt balances. Add earnings to cash as you generate them, but reduce cash by the payments on your debt. Don’t forget to account for other cash drains like working capital or capital expenditures.
The key outputs we care about are your cash balance every year (should always be above zero with some margin for error) and your annual earnings a few years out, as that defines the value of your business down the road.
Breaking the Model
We can finally assess the deal’s risk profile. As I mentioned at the top, if you assume revenue is flat or growing every year, your model will spit out numbers that make the deal look amazing.
The real question is what set of assumptions causes the model to break, by which I mean causes your cash balance to go below zero (with some margin of error, so maybe 1-2 months of debt payments at least) or leads to a very poor earnings outlook.
I start experimenting with various scenarios:
What if I lose my 1 of my top 3 customers (let’s say they represent 12% of volume and 18% of revenue) on day one? Model volume decreases by 12%, and price is adjusted down given they are paying a higher price than the rest of your customers. It’s important to differentiate between their impact on the revenue & cost drivers, not just revenue itself.
What if I’m terrible at sales, so while existing customers complete their contracts, I can’t replace them with new customers without heavy discounting? Model that volume and price go down, but not right away.
What if my staff demands retention bonuses? Model in a 10-20% increase in payroll for 1-2 years, and then 3% growth from there.
I like these scenarios to be based on real-life possibilities, not just vague “revenue down 5%” assumptions. That allows you to connect an outcome to a specific event; then you go diligence the risk of that specific event actually happening (like losing one of your top 3 customers).
As mentioned above, the outputs I’m looking at are 1) do I run out of cash and 2) where do earnings end up after a few years.
#1 tells you if the business will break in the given scenario. #2 will help you assess how much money you’ll make (assuming you didn’t break the business) in that given scenario assuming a sale in a few years off the new, lower level of earnings.
Conceptually, we all know a business with a low fixed cost base tends to have less downside risk than a business with high fixed cost base. Practically, you only see how much that matters once you build this model and start running these scenarios.
Using the Model for Diligence & Negotiations
Remember, the model is only a tool for identifying issues – it has no answers. If you put garbage assumptions in, you get garbage out. But by allowing you to create a set of scenarios to pressure test the business model, you have also figured out your guardrails – the bounds within which you need the business to operate to avoid losing your shirt.
You can now take this information directly into diligence and negotiation. If your model shows the business is more susceptible to volume declines than revenue declines (due to the nature of its variable cost base), you know where you need to press the Seller during diligence.
If the losing one of your top 3 customers does not break the business but losing one of your top salespeople does break the business, you know where you need to focus your diligence.
I’d be asking harder questions about how the Seller has retained the top salesperson in recent years; there may be massive payroll inflation in your near future depending on how that salesperson views her own value.
Further, as you think about negotiation seller notes and/or earnouts, you can maneuver negotiations to the items that really matter to you. For example, if you have a rough sense of how many units needs to be sold to breakeven from a cash perspective, you would use that to set a threshold on a seller note forfeiture clause.
From the Seller’s perspective, that can often be an easy give because they view that low level of units sales to be extremely unlikely to happen. But for me, given the information disadvantage as a Buyer, adding a bit of margin of error at that breakeven point significantly de-risks the transaction in terms of getting hit on a PG with the bank.
Another example would be negotiating deferred payments on the seller note for 12-24 months – given those are the riskiest months post-transaction and the period with the highest debt balance, providing extra cash buffer in the early period can be the difference between the model breaking or not.
Conclusion
It’s tempting as a Buyer to pencil out a flat to growing model that shows great returns. The reality is that the model will always show a great return given a reasonable multiple.
Utilize the model for a more practice purpose – assessing the risk profile of the deal, and then taking those outputs to complete diligence and negotiate the transaction.
Would love your feedback on this post as always, as well as ideas for future posts. AND — if you’re in the Seattle area and interested in an SMB community golf outing, please reply to this email!
Thanks,
Guesswork Investing