Haute Lumière

Commerce · III.05 · MMXXVI · daylight

La Bourse  /  Volume III  /  Nº III.05  /  Workbook — the Gainshare employee

A watercolour of people working at a long wooden table inside a glasshouse full of autumn trees.
Plate III.05 · Workbook — the Gainshare employeeThe Long Table.The table is the same width at both ends. It is the light that falls away. A discount rate is a claim about how fast it falls, and almost nobody who uses one has ever said out loud what theirs is made of.

WORKBOOK — THE LUMINOUS GAINSHARE EMPLOYEE

Chapter III.05 · Interest, Time, and the Discount Rate

For the person inside the mechanism. This chapter is the one that decides whether the value you create is counted at all — and, if it is counted, what it is worth to you by the time it reaches you.

WHY THIS CHAPTER IS YOURS

A gainshare pays you a share of a measured gain. Two numbers decide what you receive, and neither of them is your effort.

The first is what counts as gain. Almost every scheme in existence measures the mean: output up, cost down, waste down. Almost none measure variance. And a very large part of what skilled people actually do on the floor is variance work — the maintenance that stops the breakdown, the cross-training that stops the absence becoming a stoppage, the supplier relationship that holds in a shortage, the checklist that stops the recall. That work shows up in your employer's insurance premium and in the smoothness of their results, and it does not show up in a mean. It is the largest uncounted contribution in most schemes, and this chapter is how you get it counted.

The second is when you are paid. A gain measured in year one and paid in year four has been discounted — at somebody's rate. If it is discounted at the firm's financing cost, the firm keeps the difference between their rate and yours. That difference is not small and it is not anybody's fault. It is arithmetic, and it is negotiable once it is named.

So there are exactly two asks in this workbook, and both are specific: count the variance, and price the deferral at my rate, not yours.

PART ONE — DISCOVERY

Days 1–30: find the gain that is already there

Exercise 1.1 — Read your scheme's definition of gain (2 hours)

Find the scheme document and answer six questions in writing.

  1. What is the measured quantity, exactly? Write the formula.
  2. What is the baseline, and who signed it?
  3. Over what period is it measured, and who verifies it?
  4. What percentage is the pool, and how is it divided?
  5. When is it paid — all at once, or partly deferred? Over how long?
  6. Is any variance, volatility, downtime, loss ratio or insurance measure named anywhere in the document?

Question six will almost always answer no. That is not a complaint; it is the opening.

Exercise 1.2 — Trace one variance you personally reduce (one week)

Pick one thing that goes wrong sometimes and costs money when it does. Unplanned downtime. A quality escape. A late delivery. A safety event. A callback.

Now build three columns from whatever records exist — a maintenance log, a shift book, a complaints file, your own diary.

ColumnWhat it is
How oftenEvents per period, over as many periods as you can reach
How badCost per event, including the hours of everyone else it stops
How variableThe spread, not just the average

If you have twelve periods of data you can compute a mean and a standard deviation, and if you can compute those two numbers you can do everything in the rest of this workbook. Twelve rows in a notebook is a sufficient dataset. It is more data than most claims of this kind are ever supported by.

Exercise 1.3 — The appreciative team conversation (45 minutes)

Ask the team, out loud:

"Think of a time something could have gone badly wrong here and did not. What stopped it? Who did the thing that stopped it, and would anyone outside this room know it had happened?"

Write down the conditions, not the outcome. What you are collecting is a list of avoided events, which is exactly the list that never reaches a management account — because nothing happened, and nothing happening does not generate a document. You are building the document.

PART TWO — THE ARITHMETIC

Days 31–50: compute what you are owed

Exercise 2.1 — Your team's gain, both terms (2 hours)

Work the chapter's method on your own numbers. Here it is on a worked case so you can see the shape.

A team of twenty-four. Unplanned downtime used to cost £900,000 a year with a standard deviation of £900,000 — some years nothing much, some years a catastrophe. After the crew's maintenance and cross-training practice, it costs £640,000 a year with a standard deviation of £300,000.

  mean gain                            GBP 260,000 per year
  variance  0.81 -> 0.09 (GBP m)^2     0.72 removed
  certainty equivalent (a/2)·dVar      GBP  54,000 per year
      at a risk tolerance 1/a of GBP 6.67m
  ---------------------------------------------------------
  gain counted, mean only              GBP 260,000
  gain counted, both terms             GBP 314,000   (+20.8%)

At a pool of 20 percent divided evenly across twenty-four people:

  per head, mean only                  GBP 2,166.67
  per head, both terms                 GBP 2,616.67      (+GBP 450.00)

Four hundred and fifty pounds a head a year, for work the team is already doing, that the scheme currently values at zero. Run it on your own figures before you do anything else, because your number is the only one that will be discussed.

Exercise 2.2 — The deferral gap (90 minutes)

This is the finding most worth having and it takes twenty minutes to compute.

Suppose the scheme pays 40 percent now and 60 percent in three years. Your own discount rate — what you would genuinely accept to wait — is around 15 percent, because you have no cheap credit. The firm's financing cost is 5 percent.

  face value of your share             GBP 2,616.67
  worth to you                         GBP 2,078.97   79.5% of face
  cost to the firm                     GBP 2,402.89   91.8% of face
  ---------------------------------------------------------
  the gap                              12.4 points of face value

Twelve points of value disappear into the difference between two discount rates. Nobody took it. Nobody decided it. It is the arithmetic of a deferral written by a party with cheaper money than yours.

Now compute the fair correction. To make the deferred slice worth its face value to you, it must be grossed up by 1.15³ = 1.5209 — plus 52.1 percent on the deferred portion. That costs the firm about £706 per head in present value and restores you to par.

Write both numbers down. You are not asking for more money. You are asking for the same money, priced at the rate of the person who is waiting.

Exercise 2.3 — Your own rate, honestly (45 minutes)

Before you claim a rate of 15 percent, establish it. Three checks:

  1. What do you actually pay to borrow — overdraft, card, car finance? That is an upper bound and it is often above 20 percent.
  2. What would you accept, today, to convert £1,000 now into £X in three years? Solve r = (X/1000)^(1/3) − 1.
  3. What would you have had to give up in the last three years to not have had the money? Name it concretely.

The third check is the honest one, and it is why an employee's discount rate is genuinely higher than a firm's. It is not impatience. It is the absence of a balance sheet, and it is a fact about your circumstances rather than a fault in your character.

Exercise 2.4 — The honest negative (30 minutes)

Write the strongest case against your own claim. It exists, and you should arrive holding it.

A claim that names its own weaknesses is believed. A claim that does not is audited. Those are the only two outcomes available.

PART THREE — DESIGN

Days 51–70: make the uncounted countable

Exercise 3.1 — Build the baseline nobody built (2 weeks)

One page, and it is the most valuable document in this workbook.

  1. The metric, defined so precisely that two people compute it identically. "Downtime" is not a metric. "Unplanned stoppage minutes on line 3, logged in the shift book within the shift, costed at the standard hourly rate" is.
  2. The period. Long enough to contain normal variation — twelve periods minimum.
  3. The two statistics. Mean and standard deviation. Both, always. A baseline with only a mean has pre-decided that variance does not count.
  4. The verifier. Named. Internal audit, the risk function, or the insurance manager — the last is the strongest because they already own the loss model.
  5. Two signatures. Yours and the sponsor's, before anything changes.

Exercise 3.2 — The one-page claim (one page)

Six lines, in this order, and no more:

  1. What we changed, in one sentence.
  2. The mean before and after, with the period and the sample size.
  3. The standard deviation before and after.
  4. The certainty equivalent of the variance removed, at a stated risk tolerance, with the tolerance named and sourced.
  5. What share of it we are claiming, and what else contributed.
  6. The ask: count both terms, and price any deferral at the employee rate.

Give it to one person: whoever decides the scheme's definition of gain. Not the team meeting, not the newsletter. One person, one page, two numbers.

Exercise 3.3 — Read the reversion and the horizon (1 hour)

Two clauses in your scheme document repay reading closely.

The horizon. Over how many periods is a gain credited? A practice that reduces failures over five years but produces nothing measurable in the first twelve months will score zero under a one-year window. That is the horizon problem from the chapter, arriving in your pay packet. Ask for the window to match the asset: three years for a maintenance practice, one for a throughput improvement.

The reversion. When the scheme's investment is repaid, who keeps the ongoing benefit? If the answer is the company in perpetuity, the scheme is a loan you repaid with your own work. If the answer is the operating unit or the team, the scheme is a partnership. Ask which it is. A reversion clause is the single best test of whether a gainshare is real.

PART FOUR — DESTINY AND DELIGHT

Days 71–90: make it hold

Exercise 4.1 — Into the standing review (one conversation)

Get two numbers onto whatever is reviewed monthly: the mean of your metric and its standard deviation, side by side. Once the standard deviation is in the pack, removing it requires somebody to explain why, and nobody wants that conversation. That single column is the whole campaign, won quietly.

Exercise 4.2 — The second owner (this month)

The risk function or the insurance manager. Their mandate is already variance, they already hold the numbers, and their word carries with the audit committee in a way that yours does not yet. Bring them the baseline, not the ask, and give them the credit for the first result.

Exercise 4.3 — The written record (ongoing, ten minutes a week)

Ten minutes, once a week, in a dated notebook: what happened, what nearly happened, what stopped it. Avoided events generate no paperwork — that is their defining property and the reason they go unpaid.

A year of that notebook is a dataset. It is also, quite separately, the strongest document you will ever have in any conversation about your own value, at this employer or the next one.

Exercise 4.4 — Delight, honestly (ongoing)

There is a real pleasure in this and it is worth naming, because it is what keeps the notebook going. It is the pleasure of the moment the thing you always knew was true acquires a number — the moment somebody in a meeting says the standard deviation fell by two thirds about work you did with your hands, and the room takes it seriously.

The intuition was always right. It was simply homeless. Giving it a home is the work, and it is good work.

KNOW YOUR SCHEME — A CHECKLIST

Ten questions. Take the answers in writing.

  1. What exactly is the measured gain, as a formula?
  2. Does the definition include any variance, volatility or loss measure?
  3. What is the baseline, who signed it, and when?
  4. Over how many periods is a gain credited?
  5. What is the pool percentage, and is it of gross or net gain?
  6. How is the pool divided — evenly, by grade, by hours?
  7. What proportion is deferred, and for how long?
  8. Is the deferred portion grossed up? At whose rate?
  9. What happens at reversion — who keeps the ongoing benefit?
  10. Who verifies, and can I see their working?

Question eight is the one nobody asks, and it is worth more than questions one to seven combined.

THE CONVERSATION, SCRIPTED

Keep it to four sentences and bring the page.

"We reduced unplanned downtime from a mean of £900,000 to £640,000, and the standard deviation from £900,000 to £300,000 — that second number is a real saving and the scheme currently counts it as nothing.

At the board's own risk tolerance, the variance we removed is worth £54,000 a year, which takes the counted gain from £260,000 to £314,000 and the pool from £2,167 a head to £2,617.

Separately: the deferred sixty percent is worth 79.5 percent of face to me and costs the firm 91.8 percent, so about twelve points of value are lost in the gap between our two discount rates. Grossing the deferred slice up by 52 percent costs the firm around £706 a head in present value and puts me back at par.

I would like both changes in the scheme definition, and I have brought the baseline and the sample size."

Then stop talking. You have made two specific, costed, checkable asks, and you have named your own weakest assumption before anyone else could. That is a different kind of conversation from the one this usually is.

APPRECIATIVE QUESTIONS FOR YOUR TEAM

  1. When has something here nearly gone wrong and did not — and who made that happen?
  2. What do we do that makes our numbers steadier, and how would we prove it?
  3. Which of our practices pays off in years that are already bad?
  4. If our scheme counted variance as well as average, what would we start doing that we do not do now?
  5. What is the longest-horizon thing any of us protects, and what does it cost to keep protecting it?
  6. If we could see the company's insurance premium, what would we want to try?
  7. What would have to be true for this measurement to still be running in five years?
  8. Who outside this team would notice first if we stopped?