# Flexibility: myth and reality (what it can really absorb) "We'll just be flexible." It is the most repeated line in the FIRE movement and the least quantified. It gets used to justify a tight withdrawal rate, to wave away a simulation's bad scenarios, to end worried conversations. The instinct behind it is sound: spending you can adjust beats spending you cannot. But it turns that instinct into magical thinking, flexibility as universal absolution. The research has sorted it out, with ERN leading (Parts 23 to 25 and Part 58 of his series, among the most contrarian and the most useful, [[the-ern-series]]). **Realistic** flexibility, bounded in depth and sustainable in duration, is worth roughly 0.3 to 0.5 points of withdrawal rate. That is real and valuable. And it is three times less than what the magic phrase asks it to carry. This chapter puts numbers on all of it. Why flexibility runs into the **duration** of bad stretches, the central mechanism, and the one almost everybody misses. The six forms of flexibility ranked by what they are actually worth, starting with the most powerful, which is not a spending cut at all. An honest measure of **your own** capacity: a floor you have tested, not declared. The cost as it is lived, which shows up in the standard of living actually delivered year after year. And the condition that changes everything, flexibility **written down** and triggered by thresholds. Because intention-flexibility, the kind in the magic phrase, has a simulated value of zero. ::: cle The central mechanism: duration beats depth Instinct pictures flexibility as a shock absorber for a crash: a year or two of belt tightening while the market climbs back. Real bad stretches look nothing like that. The vintages that kill a plan run 10 to 15 years in real terms ([[market-regimes]], [[the-trinity-study]], 1966, 2000). A 15% cut holds for eighteen months on morale, for five years on effort, and never for twelve. Life keeps happening: the dentist, the furnace, the wear of going without. Realistic flexibility is **deep or long**, never both. And the scenarios that threaten a plan demand both. That is why it is worth half a point and not a point and a half. ::: ::: figure coupe-exigee-tenable The 1966 vintage, a thirty-year plan on a real US 60/40: the smallest cut that avoids ruin, as a function of how many years you hold it, for a plan at 4.5% and then for the same plan at 4.0%. The sustainable zone (15% at most, twelve years at most) is this chapter's claim, not a measurement. ::: ## Putting a number on it: what the research says Start with ERN's result (Parts 23 to 25, "Flexibility Myths"), since reproduced by everyone who has redone the arithmetic. Take a plan at 4.5 to 5% initial withdrawal, the plan flexibility is supposed to save, and run realistic flexible rules on the bad vintages. To avoid ruin, the cuts you need reach 30 to 45% of your standard of living, **held** for 10 to 20 **years**. Put another way, "the flexible 5% plan" is really a 3 to 3.5% plan that does not know it, with uncertainty and anxiety thrown in. Flexibility did not save the tight rate. It confessed, after the fact, to the prudent one. Turn it around, and **bounded** flexibility applied to a plan that was already reasonable (3.6 to 4%) does exactly its job. The simulations agree (ERN, Morningstar, and any comparison of rules that scores both failure and the variability of the standard of living, [[withdrawal-strategies-overview]]). A reversible cut of 10 to 15%, triggered inside a drawdown, brings failure probability down by 40 to 60% in relative terms, worth about 0.3 to 0.5 points of initial rate. That is flexibility's real market price, and it is a good one. Few insurance policies pay out that much for a premium this bearable. As long as you ask it for nothing more. ## The six forms, ranked by what they are worth **1. Side income (the most powerful "flexibility" of all, and it is not one).** $10k to $15k a year of temporary income during a bad stretch does what no cut can do. It shrinks the withdrawals **without** shrinking the life, for as long as you need. And it turns up more often than people expect: the accounts show FIRE retirees generating it without even looking for it ([[voices-from-real-retirees]], [[going-back-to-work]], [[pensions-and-other-income]]). Going back to work is the real release valve of a tight plan, on one honest condition. It fades with age and with distance from your old trade. It is a margin for the first fifteen years, not for the whole thirty. **2. The indexation freeze (the best pain-to-protection trade).** Skipping the inflation adjustment after a red year costs 2 to 3% of purchasing power, spread out and painless, and it disarms half of the squeeze, spending climbing with inflation while the portfolio stalls ([[fixed-inflation-adjusted-withdrawal]], [[inflation-and-withdrawal-rates]]). Call it 0.2 to 0.3 points for almost no pain. It is the first flexibility anyone should write down. **3. The bounded reversible cut (the standard).** Take 10 to 15% off the comfort level, triggered by a threshold (a drawdown past 20%, or a current withdrawal rate past some written level, [[when-to-worry]]), and lift it when the number comes back below. This is the cut-in-a-downturn rule, and the heart of the guardrails ([[guyton-klinger]], [[morningstar-guardrails]]). These are the 0.3 to 0.5 points above, provided the floor makes the cut sustainable (see below). Its gain and the freeze's do not add up, since both fire in the same years. **4. Postponing the big discretionary items.** Keep the car three more years, put off the renovation, delay the big trip: this is the natural flexibility of a budget with projects in it. It is real (5 to 10% of the budget in some years) and it is free. It also runs out. You postpone once; the second year, the postponement is a cut. **5. Temporary geographic arbitrage.** Six months a year somewhere cheaper during the crossing, the years spent inside a bear market: the accounts show how well it works (20 to 30% less spending with nothing that feels like going without, because the cost of living does the work). It is an option if you are mobile, not tied to a school year, and while your health cooperates. What counts here is the reversible version, a long stay rather than a move, with no change of tax residence. **6. The deep survival cut (never count on it).** A cut of 30 to 40%, down to bare subsistence. It exists, and the accounts show it saving plans. But it cannot be **planned**: counting it when you size the plan means planning for years of distress. It is the margin behind the margin ([[failure-probability]], the reason a simulated ruin is not a ruin as lived), not a parameter. ::: science Measuring your floor: the five-year test All of flexibility's value sits in one number that almost nobody measures honestly: the **floor**. It is the level of spending you can hold for **five years**, morale and surprises included ([[how-much-you-need]], the three-tier budget). Here is the honest method. Start from the comfort budget, line by line. Sort every line into three piles: fixed (housing, health coverage, basic food, insurance, taxes), compressible but painful (fewer outings, vacations closer to home) and compressible painlessly (the frills you have already spotted). The floor is then the fixed lines plus half the painful ones, because the accounts show that squeezing all the painful ones does not hold five years. **Then** test it: one quarter lived **at the floor**, for real, before you quit. Veterans recommend that trial quarter without exception, and it almost always moves the number 10 to 15% higher. It also turns declared flexibility into flexibility you **know**. The floor-to-comfort ratio that comes out is the plan's flexibility capacity. At 90%, you are not flexible: size the plan as a rigid one ([[fixed-inflation-adjusted-withdrawal]]). At 70 to 75%, a bounded cut is credible. Below 65%, check that the "comfort" number is not padded with frills you have not admitted to. ::: ## What it costs to live through, and the one condition Two clear-eyed points to close on. First, flexibility gets **lived**. The ruin you avoid is paid for in years of reduced spending, and the psychological asymmetry is documented: a loss weighs about double its mirror-image gain (the loss aversion of Kahneman and Tversky, 1979). Habit formation pulls the same way, because an acquired standard of living becomes the reference point you climb down from at a steep cost (Constantinides, 1990, [[lifecycle-theory]]). Cutting 12% for four years is not "nothing". It is four years of things given up, and you count them. That price deserves a look before you sign, and it can be measured. Step past the failure probability alone and read the distribution of the standard of living actually delivered, year by year, in the worst quartile of the paths. Two things show up there: how deep the cuts go, and above all how long they **last**. That is what separates one flexible rule from another ([[withdrawal-strategies-overview]]). Second, the condition that separates the flexibility that counts from the flexibility that does not exist: **writing it down**. Intention-flexibility ("we'll watch it if things go badly") has three fatal flaws. It fires too late, because the first down years get denied. It fires too small, because improvised cuts are symbolic, a canceled subscription instead of a reworked budget. And it gets improvised in fear, which produces the worst decisions in the whole repertoire ([[the-psychology-of-spending]]). Rule-flexibility (threshold written, size written, lifting written, [[when-to-worry]], [[building-your-plan]]) fires on time, in the right dose, and without a marital argument in the middle of a storm ([[couples-and-family]]). It is also the only kind a simulator can give you credit for. When it prints the failure probability of a plan "with flex", it is running a rule applied mechanically. If your real flexibility is an intention, your plan's true result is the one simulated **without** flexibility. Read that one. ::: exemple The same plan, three degrees of honesty The plan: $1.4M, comfort spending $52,000 (3.7%), 45 years. The magic-phrase version: the rate pushed to 4.3% ($60,000) "because we'll be flexible". The honest simulation (realistic bounded cuts) shows 11% central-case failure, and in the worst quartile the standard of living delivered falls to −25% for 9 years. The flexibility being invoked would have to deliver what it cannot hold. The honestly rigid version: 3.7% with no rule at all, failure at 5.5%. The written-flexibility version: 3.7%, plus an indexation freeze after a red year, plus a 12% cut whenever the current withdrawal rate goes past 4.5% (floor tested at 73% of comfort, trial quarter actually lived). Failure at 2.8%. In the worst quartile, two cut episodes, 3 to 4 years each, never below the floor they knew. Flexibility did exactly its job: it turned a good plan into a solid one. It did not turn a tight plan into a safe one, and it never will. ::: ## The essentials - Realistic flexibility is worth 0.3 to 0.5 points of withdrawal rate: real, valuable, and three times less than what the magic phrase asks it to carry. The mechanism: the bad stretches that kill a plan run 10 to 15 years, cuts hold for 3 to 5, and duration beats depth. - The six forms by value: side income (the real release valve, fading with age), the indexation freeze (the best trade of all), the bounded cut on a threshold (the standard, 0.3 to 0.5 points), postponing the big items (once), temporary geographic arbitrage (if you are mobile), the survival cut (never counted). - Your capacity is your floor, measured (the five-year test, the trial quarter actually lived) and not declared: at 90% of comfort you are not flexible; at 70 to 75% a cut is credible. - Flexibility gets lived (its price shows up in the standard of living delivered in the worst quartile, in how deep the cuts go and how long they last) and only counts once it is **written down** (threshold, size, lifting): an intention is worth zero in a simulator, and worse than zero in a storm. - It turns a good plan into a solid one, never a tight plan into a safe one: if the plan is only acceptable at maximum flexibility, the problem is the rate ([[how-much-you-need]]). --- ## Going further - Early Retirement Now, Parts 23 to 25 ("Flexibility Myths") and Part 58 ("Flexibility is Overrated"): the founding numbers ([[the-ern-series]]). - Morningstar, *The State of Retirement Income*: flexible rules judged on the life they actually deliver ([[morningstar-guardrails]]). - Why cuts cost so much to live through: Kahneman & Tversky (1979) for loss aversion, Constantinides (1990) for habit formation ([[lifecycle-theory]]). - In the simulator: the cut and its trigger ("Cut in downturns (0 = fixed rule)"), and reading the standard of living delivered ([[using-the-fire-simulator]]). - In this book: [[when-to-worry]] (the written thresholds), [[going-back-to-work]] (the most powerful form), [[guyton-klinger]] and [[floor-and-ceiling]] (the rules that frame it), [[the-psychology-of-spending]] (why intention fails).