The J-curve is the shape an angel portfolio tends to trace on a chart: it drops below what you put in, sits there for a while, then - if it works - climbs back up and past it. Down, flat, up. The path looks like the letter J. It's the single most misread feature of early-stage investing, because the part everyone sees first is the part that looks like a mistake.
It usually isn't. The dip is baked into how this asset class pays out. Understanding why is the difference between an angel who panics in year two and one who knew the trough was coming and planned around it.
What is the J-curve in angel investing?
Picture a portfolio of, say, fifteen seed cheques written over three years. Plot its total value against time. For the first stretch, the line falls. A couple of companies have already folded, a few are being carried at a markdown, and none of the survivors has had a financing round generous enough to move the needle upward. The portfolio is worth less than the cash that went into it.
Then, slowly, the curve bottoms out and turns. A breakout company raises at a higher price. Another gets acquired. The survivors start to revalue upward, and a single outsized exit can drag the whole line back above water - sometimes well above it. That's the long tail of the J.
Failure is fast. Success is slow. The J-curve is just those two speeds drawn on the same axis.
Why does a portfolio lose value before it grows?
Because the timing of bad news and good news is asymmetric.
A startup that's going to fail usually fails quickly. It runs out of runway, can't raise the next round, and shuts inside two to four years. Those write-offs cluster at the front of the portfolio's life. They're the first thing to resolve, and they resolve downward.
The companies that work take the opposite shape. Building something valuable enough to sell or float is slow - early-stage venture is generally a seven-to-ten-year-plus commitment before the meaningful exits arrive. So in the early years you've already absorbed your losers while your potential winners are still heads-down, unproven, and carried at or near cost. The gains that will eventually offset the failures simply haven't had time to show up.
This is the same engine behind the power law that governs angel returns: a minority of investments produce the bulk of the gains. The J-curve is what the power law looks like while you're waiting for it to pay off.
Paper losses aren't the same as real ones
Here's where a lot of new angels frighten themselves unnecessarily. Much of the early drop is unrealised - a valuation on a spreadsheet, not money that's left the building.
A paper loss is a mark-down. The carrying value of a holding has fallen, often because a later round priced lower, or because you've prudently written it down while you wait to see whether it survives. Nothing is fixed. The company can still recover, raise up, or exit.
A real loss is crystallised: the company is wound up, or you sell your shares below what you paid. That's the moment the loss becomes a number you can actually count - and, in the UK, the moment certain tax reliefs can come into play.
Early portfolios are heavy with paper losses precisely because the winners haven't been revalued yet. The line on the chart can look grim while the underlying position is perfectly healthy. That's the trap of reading a year-two statement as if it were a verdict.
How SEIS and EIS loss relief changes the dip
If you invest through the UK's venture capital schemes, the tax treatment of failures reshapes the downward leg of the J - it makes the trough shallower than the headline losses imply. This is one of the genuinely consequential features of investing via SEIS and EIS.
Both schemes offer loss relief to UK taxpayers who received income tax relief on the original investment. When a holding is sold at a loss or becomes worthless, you can set that loss - after deducting the income tax relief already given - against your income or your capital gains, subject to the scheme rules. Layer that on top of the upfront relief (50% income tax relief for SEIS, 30% for EIS) and the net cost of a write-off is a good deal smaller than the cash you put in.
A quick illustration of the mechanics, not a recommendation. Put £10,000 into an SEIS-qualifying company. You may have already claimed up to £5,000 in income tax relief (50%). If the company then fails, the loss eligible for relief is the £10,000 minus the relief already given - and that remaining amount can be set against income or gains under the rules. The combined effect is that the downside on a single failure is cushioned, which is exactly why the J-curve's trough is shallower for scheme investors than the raw paper losses suggest.
Two caveats worth stating plainly. The reliefs depend on holding the shares for the minimum period - three years for SEIS and EIS - and on the company keeping its qualifying status; the rules and company-side limits are detailed and change over time. And the comparison isn't universal: VCTs do not offer loss relief at all, though they carry their own reliefs (20% income tax relief, a five-year minimum hold, tax-free dividends). The exact figures, conditions and the company eligibility tests sit with HMRC; see the gov.uk guidance on tax relief for investors and the EIS company rules for the current detail.
What the shape does - and doesn't - tell you
The J-curve explains the order of events. It does not promise the upturn. Plenty of portfolios spend years underwater and never come back, because the winner that was supposed to drag the line up never materialised. A J that never lifts is just an L.
So the honest reading is this: a falling balance in the early years is normal and, on its own, tells you almost nothing about the eventual outcome. The information that matters is whether the survivors are growing, raising, and progressing - not what the aggregate mark-to-market says in month eighteen. The curve is a description of timing, not a forecast of success.
Which is why the two things that actually govern your experience of the J-curve are time horizon and diversification. Money you might need back inside a few years sits badly against an asset that spends its early life below cost. And a portfolio of three companies has no real shot at the long tail that bends the curve upward - the maths of the power law needs more cheques than that. We dig into both in how many startups make an angel portfolio and diversification in angel investing.
None of this is advice on what to do with your own capital. It's a description of how the arithmetic of early-stage portfolios behaves. Tax treatment depends on your circumstances and on rules that change; for anything specific, take FCA-regulated advice before you invest.
Frequently asked questions
What is the J-curve in angel investing?
The J-curve describes the typical shape of an early-stage portfolio's value over time. It dips first, because companies that fail tend to do so within the first few years, then rises later as the survivors grow and a small number reach a profitable exit. Plotted on a chart, that path looks like the letter J.
Why does an angel portfolio lose value before it grows?
Failure is fast and success is slow. Startups that run out of money usually do so in the first two to four years, so write-offs cluster early. The companies that work take far longer to mature and exit, so the gains that offset those losses arrive years afterwards. The order of events, not the size of the eventual return, is what creates the dip.
How long does the J-curve take to turn upward?
There is no fixed timetable, but early-stage venture is generally a seven-to-ten-year-plus game. Many angels report that their portfolios stay below cost for the first several years before maturing exits pull the line back above water. The timing depends entirely on the companies, and some portfolios never recover.
Does SEIS and EIS loss relief change the shape of the J-curve?
It softens the downward part. For UK taxpayers who received income tax relief, SEIS and EIS loss relief lets you set the loss on shares that fail - after deducting relief already given - against income or capital gains, subject to the scheme rules. That reduces the net cost of a write-off, so the trough of the curve is shallower than the headline paper losses suggest. VCT does not offer loss relief.
Is a paper loss the same as a real loss?
No. A paper loss is an unrealised mark-down: the value on the page has fallen, but you have not sold and the outcome is not yet fixed. A real loss is crystallised when a company is wound up or shares are sold below cost. Early portfolios carry a lot of paper losses precisely because the winners have not had time to be revalued upward.