The research framework
What is an inflection point?
An inflection begins when something changes inside a business that can make future earnings meaningfully better. The opportunity is to recognise that change while the share price still reflects the old business.
Story only
Management announces capex, a new product or a large target. Customers and economics have not validated it.
Usually 0% weight
Leading evidence
Orders, utilisation, cohorts, volumes or cash variables begin turning before reported earnings look clean.
The preferred hunting ground
Reported confirmation
Revenue, margin and cash flow confirm the leading evidence. Risk falls, but the price often begins recognising it.
Weight can be earned
Consensus recognition
Analyst estimates rise, the story becomes widely understood and the valuation expands.
Return now depends on delivery
Extrapolation
The market prices current growth and margins as permanent. A wonderful business can become a poor investment.
Protect against narrative risk
What must be present
Three confirmations
- Two operating confirmations
Examples include market share, utilisation, orders, repeat cohorts or product revenue. - One cash or balance-sheet confirmation
Earnings must not be manufactured through receivables, debt or accounting. - A terminal-value gate
The business must remain structurally relevant after the cycle and the current catalyst.
What does not qualify
False inflections
- Revenue financed by receivables or channel stuffing
- Margin improvement caused only by temporary raw-material deflation
- Orders without advances, capacity or payment visibility
- Peak-cycle earnings presented as a low P/E
- A new product with no customer validation
Research model methodology
From a business story to a testable investment thesis
I assess each company through four questions. Some belong here because the evidence is developing, not because every question has already been answered. The brief explains what is established, what is assumed and what is still missing.
Terminal value
Will the business remain relevant and capable of reinvesting profit long after the current catalyst is over?
Operating evidence
Have customers, utilisation, orders, volumes or cash flow moved the thesis beyond management promises?
Valuation
Do earnings scenarios offer an attractive return at the price? Assign probabilities only when there is a defensible basis, not to manufacture a precise target.
Survivable downside
If the inflection is delayed, can existing earnings, assets or the balance sheet protect against permanent loss?
Three ways to read the valuation
Probability-weighted: explicit scenario present values are combined using stated, subjective probabilities.
Conditional: a working earnings case or sensitivity, without claiming a probability-weighted answer.
Not established: the earnings, capital or funding evidence cannot yet support a reliable present value. This does not mean zero business value.
A sensitivity band changes a named assumption, it is not automatically the full downside range. The price comparison is value today divided by the latest price, minus one. Positive means the estimate is above the price, negative means it is below. Neither is an annual return forecast.
The decision rule
Evidence × asymmetry × terminal value
The result is reduced for balance-sheet, governance, cyclicality and execution risk. No single attractive number can rescue a weak thesis.
What changes the weight?
Weight rises when operating evidence improves faster than the share price. It falls when valuation runs ahead, the balance sheet weakens or a falsification trigger appears. Every change is dated and applied prospectively. A proportional redistribution can also increase a weight without improving that company’s evidence. Such changes are identified in the members’ tracker.