Reading the Hockey Stick: How to Diligence a Management Forecast

The forecast is where a deal's value gets justified — and where the most optimism quietly accumulates. Every projection is an argument: that the business will grow faster, hold margins better, or convert earnings to cash more cleanly than it has before. Good diligence does not argue with the bottom line. It tests the drivers underneath it, and asks whether the story they tell is internally consistent, grounded in history, and fundable.

Key takeaways

• A forecast is credible when it is internally consistent (the three statements tie and the balance sheet balances), grounded in history and an articulable operating narrative, and robust under stress — not merely when it looks attractive at base case.
• A balancing balance sheet proves the model is arithmetically sound. It does not prove the assumptions are right. Those are two separate tests, and conflating them is a common error.
• Sensitivity analysis flexes one variable at a time to find the levers that matter; scenario analysis flexes many variables together to model coherent states of the world. Mislabelling one as the other hides the correlation structure of the assumptions.
• The sharpest red flags are internally contradictory: growth paired with a working-capital release, margin expansion with no operational cause, capex running below depreciation while revenue scales, or a balance sheet that only balances via a forced plug.
• A forecast is not "wrong" for being ambitious. The test is whether each assumption is fundable and tells a coherent, evidenced story — and whether the model survives the downside a lender actually has to underwrite.

Why the forecast is the centre of gravity

Historical accounts tell you what a business has done. The forecast is the part of the diligence file that asserts what it will do — and it is on that assertion that price, leverage, and structure are set. A private equity buyer underwrites the base and upside cases to justify entry valuation and a return. A lender underwrites the downside: can this business still service its debt in a recession, or if rates move against it?

That asymmetry matters. The same model read by a sponsor and by a credit fund produces different anxieties. The sponsor wants to know whether the plan delivers the return; the lender wants to know whether the plan can fail gracefully. A credible forecast has to answer both, which is why it must be stress-tested rather than simply presented at the most flattering case. The lender's deliberately conservative "banking case" is not pessimism for its own sake — it is the only case that tells you whether the capital structure holds when conditions turn.

None of this requires assuming bad faith on the part of management. Most hockey-stick forecasts are sincere. The discipline of diligence is to separate sincerity from fundability.

What makes a forecast credible: the three-statement integrity check

Before you interrogate a single assumption, you check the plumbing. A forecast lives inside a three-statement integrated financial model — income statement, balance sheet, and cash flow statement linked so that a change in any driver flows automatically and consistently through all three. When it is built correctly, the three statements are mechanically linked and the balance sheet balances every period by construction.

Three bridges carry that linkage, and each is a place a model can quietly break.

The three integration bridges

Net income into equity. Net income flows into retained earnings: opening retained earnings, plus net income, less dividends, equals closing retained earnings. This is the hinge that ties the P&L to the balance sheet. If a forecast grows earnings without the corresponding build in equity, the link is severed.

Net income into cash flow. Under the indirect method, net income is the starting line of the cash flow statement. From there you add back non-cash items — depreciation, amortisation, stock-based compensation, deferred tax — and adjust for changes in working capital. The direction matters and is frequently misunderstood: a rise in receivables is a use of cash, while a rise in payables is a source. You then subtract capex and account for financing flows to arrive at the movement in cash.

Cash flow into the balance sheet. The closing cash balance becomes the cash line on the balance sheet, while the working-capital balances on the balance sheet drive the working-capital movements back on the cash flow statement. The loop closes. Wall Street Prep's walkthrough traces each of these connections in detail.

What the balancing check actually proves

When these bridges are intact, the balance sheet balances on its own — Assets = Liabilities + Equity, every period, without intervention. That is the model's internal proof of arithmetic and logical integrity. If A ≠ L+E, a linkage is broken somewhere in the chain.

Here is the trap to avoid. A balancing balance sheet proves the model is internally consistent. It does not prove the forecast is right. Those are different claims. A model can be flawlessly constructed and still rest on assumptions that have no basis in the business. The integrity check tells you the arithmetic holds; it tells you nothing about whether revenue will actually double.

The corollary is the first real red flag. If the balance sheet only balances because someone has inserted a forced "plug" — a hardcoded number that makes the equation close — the linkage is broken and the plug is hiding it. A correctly built model never needs one.

Sensitivity is not scenario: a precision point

These two terms are routinely used interchangeably, and the conflation is not pedantic — it changes what the analysis can tell you.

Sensitivity analysis changes one input variable at a time, holding everything else constant, to measure how responsive an output is to that single driver. How does IRR or equity value move as revenue growth flexes by plus or minus two percent? The purpose is model validation: isolating which levers actually matter and which barely move the outcome. In practice these are usually one- or two-variable data tables.

Scenario analysis changes multiple variables simultaneously, in internally coherent bundles, to model whole states of the world — typically a base case, an upside or bull case, and a downside or bear case, plus the lender's banking case. The point is to tell a coherent story about how a combination of conditions plays out together. A recession is not "lower revenue, all else equal." It is lower volume and margin compression and higher financing cost, all at once and all correlated. CFI's comparison of the two methods sets out the distinction cleanly.

The key difference: sensitivity isolates the effect of one variable; scenario captures the joint effect of many. A common modelling sin is to relabel a genuine multi-variable scenario as a "sensitivity." Doing so strips out the correlation structure of the assumptions — it presents a coordinated downturn as though its components were independent, which understates how bad a real downside looks. When a model offers only single-variable sensitivities and no coherent downside scenario, that absence is itself a finding.

What diligence scrutinises: test the drivers, not the bottom line

With the plumbing verified, diligence moves to the assumptions. The discipline is always the same: trace each line of the forecast back to something demonstrable in the operations, and check whether the assumptions are consistent with one another.

Hockey-stick growth

The signature pattern is a forecast growth rate substantially above what the business has historically achieved, with no demonstrable catalyst to explain the inflection. Growth has to come from somewhere — new capacity, a signed pipeline, specific pricing actions, a market expansion. The classic test is internal coherence: a model that shows revenue doubling but cannot explain the corresponding increase in marketing spend, headcount, or customer-acquisition cost is not telling a coherent story. The cost base has to grow with the revenue, and if it does not, the forecast is asserting an efficiency the business has never demonstrated.

Buyers respond to this by validating the sales methodology, the claimed market catalysts, and the documented evidence behind the pipeline. As practitioners who handle hockey-stick projections in M&A note, unsubstantiated forecasts get repriced sharply — or they kill the deal outright. The point is not that ambition is disqualifying. It is that ambition without a fundable, evidenced catalyst is just a number.

Margin expansion versus history

A close cousin is the forecast that bakes in gross or EBITDA margins materially above the historical trend without an identified, fundable lever. Margins can improve — through scale economies, a shift in product or customer mix, or a specific, costed cost programme. What should give a reviewer pause is a margin ramp with no operational cause attached to it. If the model assumes margins climb but cannot point to the mechanism, the expansion is an assumption masquerading as a result.

Working-capital assumptions

This is where some of the most subtle optimism hides. Watch for forecasts that quietly improve days sales outstanding, days inventory outstanding, or days payable outstanding — releasing cash from working capital to flatter projected free cash flow.

The tell is internal contradiction. High growth normally consumes working capital: more sales mean more receivables and more inventory to fund. A forecast that pairs aggressive growth with a working-capital release is asserting two things that do not usually coexist. Either the growth is overstated or the working-capital assumption is, and a reviewer is entitled to ask which.

Capex versus depreciation

If forecast capex runs below depreciation while revenue grows, the business is implicitly shrinking its asset base while expanding its output. That is usually unsustainable, and it is often a sign of deferred maintenance capex flattering near-term cash. Over time, maintenance capex should at least track depreciation; a forecast that lets capex drift below it is borrowing cash from the future condition of the asset base.

Cash conversion and earnings quality

The summary question is whether forecast EBITDA actually converts to cash. A widening gap between EBITDA and operating cash flow — driven by a working-capital build, capitalised costs, or aggressive revenue recognition — signals low earnings quality. Reported profit that does not turn into cash is the recurring theme behind most of the individual red flags above, and it is worth tracking as a metric in its own right across the forecast period.

EBITDA add-back quality

Finally, scrutinise the normalisation adjustments. Persistent upward "normalisation" — one-offs that somehow recur every year, owner-compensation add-backs, pro-forma run-rate synergies booked as though they were historical — masks real costs and inflates the earnings base the whole forecast is built on. A genuine one-time item appears once. An adjustment that reappears annually is a cost, not an exception.

The red flags, consolidated

Pulled together, the patterns that warrant escalation are:

• Hockey-stick revenue with no fundable, documented catalyst.
• Margin expansion unsupported by an operational cause.
• A working-capital release assumed alongside growth — internally inconsistent.
• Cash conversion deteriorating versus history.
• Forecast capex below depreciation while the business scales.
• Recurring "one-time" add-backs in the normalised earnings base.
• Revenue-recognition aggressiveness widening the EBITDA-to-cash gap.
• A balance sheet that only balances via a forced plug.
• Customer or supplier concentration left unaddressed in the forecast.
• Reliance on one or two individuals for the numbers — a governance and integrity flag as much as a modelling one.
• A model that cannot tell a coherent story linking revenue to its cost base.

What unites the most serious of these is internal contradiction. A forecast that is merely ambitious can still be defended with evidence. A forecast that is internally inconsistent — growth without the cost base, free cash flow conjured from a working-capital release, output expanding while the asset base shrinks — cannot be, because its own assumptions disagree with each other.

A note on UK and US targets

The three-statement linkage logic is framework-agnostic. Net income flowing to retained earnings, cash flow tying to the balance sheet, and the balance sheet balancing all hold identically under IFRS, UK GAAP/FRS 102, and US GAAP. A model built on one framework is not structurally less sound than one built on another.

The biggest model-level divergence is lease accounting. IFRS 16 uses a single on-balance-sheet model for leases, whereas US GAAP's ASC 842 retains a dual model in which operating leases keep a single straight-line expense above EBITDA. That difference affects forecast EBITDA, depreciation, interest, and net debt — and it must be normalised before a UK target and a US target can be compared on a like-for-like basis. Reviewing cross-border portfolios without adjusting for it produces apples-to-oranges multiples and a misleading read on leverage. Advisory practices that run cross-border deal diligence treat this normalisation as routine, and so should any reviewer comparing the two.

Frequently asked questions

Does a balancing balance sheet mean the forecast is reliable?

No. A balancing balance sheet proves only that the model is internally consistent — the arithmetic and the linkages hold. It says nothing about whether the underlying assumptions are credible. A perfectly constructed model can rest on a revenue assumption with no basis in the business. Integrity and credibility are two separate tests, and you have to run both. The one thing a balancing check does flag is a forced plug used to make it balance, which is a sign the linkage is actually broken.

What is the practical difference between sensitivity and scenario analysis?

Sensitivity analysis changes one variable at a time, holding everything else constant, to find which drivers most affect the output — it validates the model and isolates the key levers. Scenario analysis changes many variables together in coherent bundles to model whole states of the world, such as a recession in which volume falls, margins compress, and financing costs rise simultaneously. Sensitivity gives you the isolated effect of one input; scenario gives you the joint effect of several. Relabelling a multi-variable scenario as a "sensitivity" hides the way the assumptions move together.

Why do lenders and equity investors read the same forecast differently?

A private equity buyer underwrites the base and upside cases to justify the entry valuation and target return. A lender underwrites the downside — whether the business can still service its debt in a recession or at higher rates — and builds a deliberately conservative banking case to test it. The forecast has to satisfy both, which is exactly why a credible model must be stress-tested rather than presented only at its most flattering point.

Is an ambitious forecast automatically a red flag?

No, and treating it as one is its own error. A forecast is not wrong merely for being optimistic. The real test is whether the growth, margin, working-capital, and capex assumptions are fundable and whether together they tell a coherent, evidenced story. Ambition backed by a signed pipeline, identified capacity, and a cost base that scales with revenue is defensible. Ambition with none of that attached is just a number on a slide.

How DiligenceForge approaches forecast diligence

DiligenceForge helps lenders and investors test management projections against the standards above — checking that the three statements tie, separating isolated sensitivities from coherent downside scenarios, and surfacing the internally contradictory assumptions that distinguish an ambitious plan from an unfundable one. The aim is a faster, more defensible read on whether a forecast survives contact with the downside a lender actually has to underwrite. To see how it works on your own deals, request access.