Haute Lumière

Commerce · V.11 · MMXXVI · daylight

La Bourse  /  Volume V  /  Nº V.11  /  Workbook — the executive

WORKBOOK — THE CORPORATE EXECUTIVE

Chapter V.11 · When Flourishing and Output Diverge

For the person with a P&L, a signature limit, a board and a quarter. This workbook uses the language of the firm without apology, because your own numbers already contain the finding — they have simply never been put on the same page as each other.

---

THE PREMISE, STATED COMMERCIALLY

You are being asked to do one thing: stop letting a flow measure answer a question only a stock measure can answer.

Your P&L reports flows. Every one of them can be raised for a period by drawing down a stock the P&L does not carry — hours, attention, skill, relationship, maintenance, trained capacity. That withdrawal does not appear as a withdrawal. It appears as performance, and it is indistinguishable from performance until about four to six quarters later, when the stock cannot fund another one.

This is not an ethical observation. It is the same accounting error Volume I named in the fixed asset register — an appreciating asset on a declining schedule — arriving through the payroll instead of through capex. Neither error shows up as an error. Both show up as a good quarter.

The fix is not a new philosophy, a new index or a culture programme. It is two columns instead of one, a threshold written before the first divergence, and an enquiry owned by somebody who owns neither number. The whole thing can be in your pack inside ninety days and it costs the price of a verification fee.

---

PART ONE — DISCOVERY

Days 1–30

Exercise 1.1 — The divergence audit (one week, with your controller)

Do not commission a study. Sit with your controller and go back three years.

Pull, by business unit and by quarter:

Now plot them against each other and mark every quarter where a flow measure rose while a stock measure fell by anything material. You are not looking for a culprit. You are looking for how often it has already happened without anyone noticing, and the answer is almost always more often than the room expects.

Output: a list of divergent quarters, with the unit and the pair.

Exercise 1.2 — The one that already went right (half a day)

Now the appreciative half, and do it before anything else, because it changes what the rest of this is.

Find the quarter in the last three years where a flow measure and a stock measure both improved. There will be one. Interview whoever ran it and ask:

"What were the conditions in that quarter? What did you do that you do not normally do? And what would it take to have more of that?"

Take notes on conditions, not outcomes. What you are assembling is evidence that this organisation already knows how to do this — a materially easier argument than the claim that it must learn.

Exercise 1.3 — The honest positive, for your board (2 hours)

You will be asked whether any of this pays. It does, and the number is public.

Edmans (2011), Journal of Financial Economics: a portfolio of the "100 Best Companies to Work For in America", 1984–2009, returned a four-factor alpha of about 3.5 percent a year, roughly 2.1 percent industry-adjusted — 2.45× and 1.72× cumulative over twenty-six years.

Present it with the honest reading attached, because somebody in the room will get there first otherwise: an alpha is evidence of mispricing. Employee satisfaction paid because the conventional measures could not see it. Two consequences, both useful to you. Measuring it is an information advantage, not a cost centre. And the advantage decays as the measure spreads — which is an argument for building it this year, not an argument that it pays for ever.

---

PART TWO — THE ARITHMETIC

Days 31–45

Exercise 2.1 — Compute your own sigmas (half a day)

The single most useful number in this workbook, and almost no executive has it.

For each of your two chosen measures, take three years of period-to-period changes and compute the standard deviation of the change. Not of the level — of the change.

Most people are astonished by how wide it is. That width is why a rule keyed to the sign of a movement is worthless, and it is why your threshold needs a materiality gate.

Exercise 2.2 — The trigger arithmetic (1 hour)

Learn this well enough to produce it in a meeting, because a credit officer will ask.

Two series doing nothing at all:

  sign alone      P(opposite in one period)      = 0.50
                  P(twice running)   0.5²        = 0.25      = 25 %
  with a gate     one-tailed normal, |move|>1sd  = 0.1587
                  P(opposite AND both material)  = 2 × 0.1587² = 0.0503
                  P(twice running)   0.0503²     = 0.00253   = 0.25 %

A trigger on sign alone fires a quarter of all periods on pure noise. Gated at one standard deviation it fires about once in four hundred — an improvement of roughly 99× for one sentence of drafting.

One standard deviation is a policy choice; 1.5 or 2 are defensible and will fire less. Sign alone is not defensible, and now you can say so with a number.

Exercise 2.3 — Cost your own retention effect (half a day)

The chapter's worked case, with your figures substituted:

  people not replaced   = staff × fall in voluntary turnover
  value                 = people × salary × fully loaded replacement factor

Worked: 600 staff, a 3-point fall in turnover, £45,000 salary, a replacement factor of 0.75 — the reported range is 0.5 to 2.0 times salary, so use your own and cite the range as a sanity check. That is 18 people and £607,500 a year.

State your replacement factor and its source in the paper. Executives who monetise everything are discounted entirely; an executive who names the range, picks the conservative end of it and says so is believed.

Exercise 2.4 — The permanent divergences in your own operation (one week)

The honest part, and the part that earns you the right to the rest.

Go and find where your business runs on a divergence that no design removes, and name who carries it.

• Shift work. A plant runs 168 hours a week against a day shift's 40 — a 4.2× multiple on capital utilisation you cannot obtain any other way. The IARC classifies circadian-disrupting shift work as probably carcinogenic, Group 2A (2007, reaffirmed 2019). What premium do you pay, and on what basis was it set? For most firms the honest answer is convention.
• Care. How many of your people are carrying unpaid care alongside the job? Globally the ILO (2018) counts 16.4 billion hours a day — 2.05 billion full-time equivalents — with women doing 76.2 percent. Your own figure is obtainable and almost certainly unobtained.
• Commuting. Stutzer and Frey (2008) find long commutes are not compensated: an hour each way needs about 40 percent more income to leave a person as well off — £18,000 a year on a £45,000 salary, a shadow price of £39 per commuting hour that nobody pays. That is a real cost your location decision imposes and does not carry.

Write one page naming each divergence in your operation and who bears it. A board paper that names the party on the losing side is trusted with everything else in the paper. One that calls it a trade-off and moves on is not.

---

PART THREE — DESIGN

Days 46–60: the instrument

Exercise 3.1 — Choose the pair (one meeting)

One flow, one stock, capable of disagreeing. If they are mechanically linked — revenue per head against headcount — the pair can never diverge and therefore can never inform. Pick again.

The stock measure must be behavioural, not self-reported. Voluntary turnover, absence, internal fill rate, hours worked against contracted. A survey can sit beside these; it cannot be the covenant KPI, for the reasons in Exercise 3.3.

Exercise 3.2 — Draft the facility (one week, with treasury)

A paired-KPI facility with a divergence standstill. Mechanically a sustainability-linked loan, which your treasury already understands.

The number that decides it, on the front page:

  value of the retention effect  £607,500   /   ratchet  £40,000   =  15.2 x
  net value  £582,500            /   verification  £25,000        =  23.3 x

On a £40 million facility a 10 basis point ratchet is worth £40,000 a year. The ratchet is worth a fifteenth of the behaviour it prices. The margin was never the incentive. What the facility buys is an externally audited paired series the organisation cannot quietly stop producing, and that returns 23 times its own verification cost in one line item.

Price the measurement, not the margin. Say that sentence in the meeting.

Exercise 3.3 — Make your stock measure unfalsifiable-by-flattery (one week)

Three conditions turn a flourishing measure into an artefact. Close all three.

1. Self-reported and not anonymous. A survey whose results reach a line manager measures the relationship with the line manager. Close it: third-party administration; publish the response rate beside the score every single time; and treat a rising score with a falling response rate as a falling score.
2. No anchor, running long enough to adapt. Hedonic adaptation is real (Brickman, Coates and Janoff-Bulman, 1978) and a pure self-report series drifts toward its own mean. Close it: pair it with a behavioural measure that cannot adapt.
3. In somebody's bonus with no denominator. Any measure that pays somebody and states no denominator will improve. Close it: publish what the measure did not look at, in the same table.

And the limit that protects your people. Lucas, Clark, Georgellis and Diener (2004) found life satisfaction does not return to baseline after unemployment. Set-point theory is wrong in the case where being wrong matters most. Adaptation is never a reason to discount a fall, and any executive who uses it that way should be corrected in the meeting.

Exercise 3.4 — Name the enquiry owner (one conversation)

Somebody who owns neither number. Internal audit is the natural home and is usually free. If the flow's owner runs the enquiry, the finding will be that the stock measure is soft. If the stock's owner runs it, the finding will be that the flow was borrowed. Neither is an enquiry.

Terms of reference: one question — what was the flow drawn from? — three weeks, and a written answer that goes in the pack whether or not it flatters anyone. An enquiry with no publication requirement will always conclude that nothing is wrong.

---

PART FOUR — DESTINY AND DELIGHT

Days 61–90

Exercise 4.1 — Both columns, one page, one pack (one conversation)

The mechanism is a layout decision before it is anything else. A stock measure in a people deck and a flow measure in the finance pack will never be seen diverging, because nobody holds both at once.

This is usually a five-minute conversation with whoever owns the pack — a conversation nobody has because it feels too small to ask for.

Exercise 4.2 — Publish the threshold cold (before day 60)

Write it, date it, circulate it, before your first divergence. A threshold agreed after a number has moved is a negotiation about that number.

Then let it fire on you once without adjusting it. That is the moment it becomes real, and everyone in the building will be watching to see whether you move it.

Exercise 4.3 — Put the enquiry in the compensation conversation (one cycle)

Not the metric — the enquiry. The committee's standing question becomes: what was this quarter's outperformance drawn from?, asked with the stock measures for the same quarter beside it.

A quarter bought out of a stock is still a good quarter. It is recorded as a withdrawal, and the board knows the balance. That single sentence is the whole governance change.

Exercise 4.4 — Take the clause to the bank (day 76–90)

Your existing relationship bank, at the next refinancing, not as a new facility. You are asking for a second KPI and a standstill clause, not a new product.

Bring the trigger arithmetic. The credit officer's first objection will be that the clause will fire constantly, and the answer is 0.253 percent.

Exercise 4.5 — Delight, for a firm (ongoing)

Delight is the adoption mechanism, not the reward. In a firm that means: the divergence meeting is short and has a fixed question; the pack is beautiful and fits on a page; the enquiry has a name and a date rather than a committee; and when the answer is nothing was drawn from anything, the series just moved — which it will be about half the time — nobody is made to apologise for it.

Every one of those is an adoption decision disguised as an aesthetic one.

---

THE FAILURE MODES, NAMED

1. The pair was chosen to agree. Mechanically linked measures cannot diverge, so they cannot inform. Prevention: pick a pair that can disagree, and test it against three years of history.
2. The threshold was set on sign alone. Fires 25 percent of periods, everyone learns to ignore it, dead within a year. Prevention: the one-sigma gate.
3. The threshold was written after the first divergence. It is then a negotiation about that divergence. Prevention: publish it cold, day 45.
4. The enquiry was run by one of the two number-owners. Prevention: internal audit, and terms of reference in writing.
5. The stock measure is a non-anonymous survey. It will rise while conditions fall, and the response rate will tell you so if you publish it. Prevention: the three conditions in Exercise 3.3.
6. The divergence was real and permanent and nobody named who was carrying it. Prevention: Exercise 2.4, in the board paper, by name.
7. Somebody read this chapter as saying output does not matter. It does. Output pays the wages, funds the training and clears the covenant. The claim is narrower and harder: a measure that cannot tell income from a withdrawal should never be the only number in the room.

---

THE NINETY DAYS ON ONE PAGE

---

BOARD PAPER TEMPLATE

Title. One line, commercial. "Paired-KPI facility: £40m, ±10bp, retention effect worth 15× the ratchet." Not "Wellbeing initiative."

1. The number. Retention value over ratchet, and net value over verification cost, in the first paragraph.
2. What we found. The divergence audit: how many quarters in the last three years showed a flow rising while a stock fell, by unit.
3. The one that went right. The quarter where both improved, and the conditions that produced it.
4. The pair, and why it can disagree.
5. The threshold, with the trigger arithmetic: 25 percent ungated against 0.253 percent gated.
6. The instrument. Two KPIs, ±10 bp earned only on both, the standstill, verification, term, security unchanged.
7. The permanent divergences. Named, with who bears each. This paragraph is why you will be believed.
8. What would make this fail. Three honestly, from the list above.
9. At ten times. One paragraph: what the next decision looks like.

Length: two pages. If it is longer, you have not decided what matters.

---

APPRECIATIVE QUESTIONS FOR YOUR LEADERSHIP TEAM

1. When did a flow measure and a stock measure last tell us different things, and what did we learn by taking the disagreement seriously?
2. Which quarter in the last three years improved both — and what made it possible?
3. Where are we currently reporting a withdrawal from a stock as operating income, and how would we know?
4. Which two measures would we choose that are genuinely capable of disagreeing?
5. What threshold would we be willing to publish before we knew which way it would fire on us first?
6. Who here could run an enquiry into a divergence without owning either number, and what would we have to give them?
7. What is our fully loaded replacement cost per leaver, and who has ever computed it?
8. On what basis was our shift premium set, and when was that last examined?
9. How many of our people are carrying unpaid care alongside this job, and have we ever asked?
10. What does our location decision cost the people who commute to it, and what would it take to carry that number?
11. If a lender priced our capacity as well as our output, what would we attempt that we will not attempt today?
12. What is the first sign we would see if one of our two numbers were quietly becoming decorative, and who would notice first?