Skip to main content
All posts

Stress-Testing Purchasing Power: A Framework for Long-Term Capital Preservation

InvestWise4UNovember 19, 20265 min read

A single projection answers the question "what happens if everything goes as assumed?" Long horizons almost never go as assumed. Inflation arrives in bursts, markets fall at inconvenient moments, and money is needed when it is needed, not when prices are good.

Stress-testing replaces the single projection with a small set of deliberately unpleasant scenarios, so that the plan is judged on how it fails rather than on how it performs in the average case. This guide sets out three tests and a way to build protection against each.

Test 1: an inflation spike

The usual projection assumes a constant inflation rate. Real economies tend to produce long stretches of moderate inflation interrupted by short episodes of much higher prices. Because price rises compound and are not reversed afterwards, a few bad years permanently lower the purchasing power of everything held in nominal terms.

Take $100,000 earning a steady 6% nominal a year for 10 years, and compare three inflation paths:

ScenarioInflation pathPrice level after 10 yearsReal value of $179,085
Baseline3% every year1.344$133,256
Spike9% for three years, then 3%1.593$112,439
Persistent5% every year1.629$109,943

The nominal balance is identical in all three: $179,085. The spike scenario averages under 5% a year, yet it costs nearly as much purchasing power as a full decade at 5%, because three years at 9% alone raise the price level by 29.5% and nothing brings it back down.

Two practical conclusions follow. First, a plan built on the baseline has no margin; the spike costs about 16% of the real outcome. Second, the assets that hurt most in a spike are those with fixed nominal payments, such as long-term fixed deposits or long bonds bought before the spike, because their payments cannot rise with prices. Assets whose income or value can reprice, such as inflation-linked bonds, short-term instruments that roll over at new rates, and, over longer periods, equities and real estate, have historically provided more protection, though none reliably in every episode.

Extreme cases, where monthly inflation runs at double digits, are outside what any portfolio arithmetic can manage; the defences there are holding assets in another currency or in real goods. For most savers, the realistic scenario to test is the spike above, not hyperinflation.

Test 2: the order of returns

Average returns describe a portfolio badly once money is being withdrawn. Consider two sequences over ten years. Sequence A starts with two bad years (−20%, then −10%) followed by eight years at +12%. Sequence B has the same returns in reverse order. Both average 6.6% a year.

With no money going in or out, the order does not matter: multiplication is commutative, and $500,000 becomes $891,347 under either sequence.

Now withdraw $30,000 at the end of each year:

SequenceAverage returnFinal balance with $30,000 a year withdrawn
A: losses first6.6%$381,226
B: losses last6.6%$571,673

Same returns, same withdrawals, and a difference of about $190,000. In sequence A, the withdrawals in the first two years sell assets at depressed prices, and those units are not there to benefit from the recovery. This is sequence-of-returns risk, and it is the dominant risk in the years just before and after someone starts drawing on savings.

The test is simple to run on any plan: put the worst years at the start of the withdrawal period and see whether the plan survives.

Test 3: a forced sale

The third test asks what happens if you need money at the worst time. Suppose a portfolio of $500,000 with spending needs of $40,000 a year. If everything is invested in growth assets and they fall 25%, the portfolio is worth $375,000, and the next year's spending must be raised by selling into the fall.

Compare a structure that holds two years of spending, $80,000, in cash or short-term instruments, with $420,000 in growth assets:

  • The growth portion falls 25%, to $315,000. It needs a 33.3% gain to recover.
  • Spending for the next two years comes from the cash portion. Nothing is sold at the low.
  • Had $40,000 been raised from the depressed growth assets instead, it would have required selling 12.7% of them, against 9.5% at the pre-fall value.

The cash portion earns less in normal years, and that is the cost of the protection. What it buys is time: two years in which the growth assets can recover without being sold.

Building the defences

The three tests point to a layered structure, often called a bucket approach. The proportions are personal; the logic is general.

  1. Emergency reserve. Three to six months of essential spending, instantly accessible. Its job is to absorb job loss or an unexpected bill without touching anything else. Return is not its purpose.
  2. Near-term spending. Money needed within roughly one to three years, held in short-term instruments: deposits, Treasury bills, money market funds. This is the layer that defuses Tests 2 and 3. Its yield should at least be checked against inflation after tax; the break-even nominal yield is π / (1 − t), which at 4% inflation and a 25% tax rate is 5.33%.
  3. Long-term growth. Everything not needed for several years, invested for real growth. This layer carries the market risk, and it can do so precisely because the two layers above mean it is never a forced seller.

The structure needs maintenance. In good years, gains from the growth layer refill the near-term layer. In bad years, the near-term layer is drawn down and the growth layer is left alone.

Running your own stress test

You need three numbers: what you have, what you add or withdraw each year, and the real outcome you need. Then:

  1. Project the baseline in today's money, using a realistic inflation rate.
  2. Rerun it with inflation three points higher for three years, or simply several points higher throughout.
  3. If you are drawing on the money, place the worst plausible years at the start.
  4. Check that the near-term layer covers the spending you could not delay.

The inflation-adjusted wealth simulator shows the nominal and real balance side by side; running it twice, at your baseline inflation rate and at a higher one, covers steps 1 and 2. The compound growth calculator is useful for checking the effect of different contribution levels. Neither tool forecasts; both show you the arithmetic of an assumption you choose.

A plan that only works in the average case is not robust. A plan that survives these three tests usually is, and the protection it carries costs a known, modest amount of return in the years when nothing goes wrong.


About this guide. Published by InvestWise4U, an independent site run by Ayaz Mirza (LinkedIn). Every figure above is a worked example calculated for this article, not a forecast and not advice for your situation. How the site sources and labels its numbers is set out in the editorial policy and the financial disclaimer.

Keep reading

Ready to see the actual numbers?

InvestWise4U ranks the investment options available in your country against your amount, timeframe and risk tolerance — with real yields, fees and lock-in periods spelled out.