How Many Platforms Is Enough, and Why Loan Count Is the Wrong Answer
Spreading across loans handles borrower default and does nothing about platform failure. 2020 proved that at scale.
- Loan count diversifies borrower risk, not platform risk. 500 loans on one platform all sit behind one locked door.
- Our stances: 2 platforms at 7-9% realised, 4 at 9-11%, 6 at 11-14%. These are yield bands, not rigid targets.
- Size each platform so that one going to zero is survivable. That means capping any single platform at 20-30% of your P2P capital.
- Spreading further than 6 platforms adds complexity without meaningful risk reduction, unless you are running a seven-figure portfolio.
- Capital at risk. No compensation scheme covers borrower defaults. Platform failures have frozen investor funds for months or years.
What loan count actually protects against
Spreading EUR 10,000 across 200 loans on one platform reduces the damage from any single borrower defaulting. If one loan goes bad, you lose EUR 50. If the platform's buyback promise holds, you lose nothing. If the loan sits in a pool with collateral, the recovery might bring back EUR 30 of the EUR 50. That is borrower diversification, and it works exactly as advertised.
What it does not protect against is the platform itself stopping payments. In 2020, several European platforms with thousands of loans per investor froze withdrawals or moved into managed wind-downs. Investors with 500 perfectly diversified loans found themselves locked out of all 500 at the same time. The loan spread helped nothing. The platform was the single point of failure.
That distinction is the reason P2P diversification means platform count first, and loan count second. You diversify across platforms to survive a platform failure. You diversify across loans within each platform to reduce the noise from individual borrower defaults.
The 2-4-6 framework for platform count
Our framework ties platform count to your target yield band. Lower yields come from platforms with stronger track records and tighter supervision. Higher yields come from platforms with weaker oversight or shorter histories. To reach a higher realised return without unacceptable concentration, you need more platforms.
For a 7-9% realised return band, two platforms is enough. Pick two names rated 70% match or higher, both with MiFID II or ECSP licences, and split your capital 50-50. Mintos and Capitalia give you that profile. Mintos scores 80% match and holds a MiFID II licence from Latvijas Banka, covering up to EUR 20,000 in claims against the firm if it fails. Capitalia scores 80% match and operates under an ECSP licence, also from Latvijas Banka, and holds an InvestEU guarantee worth EUR 15 million. One platform failure loses half your P2P capital, which is survivable if P2P is a modest slice of your total wealth.
For a 9-11% realised return band, four platforms spreads the risk more sensibly. At this yield level, you are mixing 70-80% platforms with a 65-70% name or two. Four platforms at 25% each means one failure costs you a quarter of the P2P allocation. Mintos, Capitalia, InRento and Maclear give you that mix. Maclear scores 65% match, sits under a Swiss SRO registration that covers anti-money-laundering supervision only and offers no compensation fund, and has covered its one default in full from the CEO's own funds. That SRO registration is lighter than EU licensing, which is why Maclear sits lower in the ranking despite a clean payment record.
For an 11-14% realised return band, six platforms is the minimum to avoid dangerous concentration. You are now holding platforms in the 60-70% band, some with no secondary market, some with concentrated originator networks. Six platforms at 16-17% each means one failure costs you that slice and the other five continue. Indemo, Profitus, Crowdpear, PeerBerry, Maclear and InSoil give you six independently operated entities across four countries, each with a different loan type and a different regulatory setup. One platform going to zero is painful but not catastrophic.
Why going beyond six platforms rarely makes sense
Spreading across seven, eight or ten platforms adds work without adding much safety. Once you hold six platforms, each capped at 20% or less, the damage from any single failure is already contained. Adding a seventh platform and rebalancing to 14% each reduces the single-platform loss from 16% to 14%. That two-point improvement costs you another set of auto-invest rules, another tax form, another login to remember, and another chunk of time watching for platform news.
The exceptions are large portfolios and professional allocators. If you are running EUR 100,000 or more in P2P, spreading across eight to ten platforms keeps any single position under EUR 15,000, which can matter for liquidity and for keeping positions under compensation fund thresholds where those apply. But for retail investors with EUR 5,000 to EUR 30,000 in P2P, six platforms is the practical ceiling.
Sizing positions so one failure is survivable
Platform count means nothing if you size the positions badly. Holding six platforms with 50% of your capital on one and 10% on each of the other five is not diversification. The dominant position is still a single point of failure.
The rule is simple: size each platform so that it going to zero leaves you able to continue. For most people, that means capping any single platform at 20-30% of the P2P allocation. If you hold EUR 10,000 in P2P and split it across four platforms at EUR 2,500 each, one platform freezing costs you EUR 2,500. Painful, but you still have EUR 7,500 working. If you put EUR 7,000 on one platform and EUR 1,000 on each of three others, the big position going down costs you 70% of the portfolio. That is not diversification.
The portfolio mix tool runs this calculation for you. Enter your target yield band and your total P2P capital, and it shows you how many platforms to use and how much to put on each. The sizing assumes equal weight across platforms unless you deliberately overweight a higher-rated name, in which case you rebalance the others downward to keep the total at 100%.
What yield band you can actually reach with N platforms
| Platform count | Realistic yield band | Risk profile | Example mix |
|---|---|---|---|
| 2 | 7-9% | Both 70%+ match, both EU-licensed | Mintos 50%, Capitalia 50% |
| 4 | 9-11% | Mix of 70-80% and one 65-70% | Mintos 25%, Capitalia 25%, InRento 25%, Maclear 25% |
| 6 | 11-14% | 60-70% platforms, some without secondary markets | Indemo 17%, Profitus 17%, Crowdpear 17%, PeerBerry 17%, Maclear 16%, InSoil 16% |
These are realised yield bands, not advertised. Advertised yields run 1 to 4 percentage points higher than what investors actually receive, once you account for defaults, buyback delays, cash drag and early repayments. The bands above assume you hold each platform for at least 18 months and reinvest distributions as they arrive.
When one platform is acceptable
Single-platform portfolios make sense in exactly two situations. The first is when you are testing P2P with a small amount you can afford to lose entirely. Putting EUR 500 on one platform to see how the mechanics work is fine. You learn whether you can tolerate the illiquidity, whether the tax reporting annoys you, and whether watching the portfolio adds stress or boredom. One platform going bad costs you EUR 500, which is the cost of the education.
The second situation is when you are using a single high-rated platform as a tactical cash position. Some investors park EUR 2,000 to EUR 5,000 on Mintos or Capitalia with auto-invest turned on, earning 9-10% while they decide where to deploy the capital next. The P2P position is temporary, the platform is 80% match, and the amount is small enough that a freeze would be inconvenient but not damaging. That is a deliberate choice, not an accident.
Outside those two cases, single-platform portfolios concentrate all your platform risk in one entity. That entity can freeze withdrawals, lose its licence, face a liquidity crisis, or simply wind down operations. When it happens, 100% of your P2P capital is stuck. Two platforms cuts that exposure in half. Four platforms cuts it to a quarter. The work of managing multiple platforms is modest compared with the cost of a total lockout.
Common questions about platform count
No. Loan count diversifies borrower default risk, not platform failure. If the platform freezes withdrawals, all 500 loans sit behind the same locked door. 2020 saw platforms with thousands of loans per investor go into workout or wind-down. The loan spread helped nothing.
Borrower diversification matters, and 200 to 500 loans per platform is a sensible target. But it does nothing about the platform itself becoming the point of failure. That requires holding multiple platforms.
Two platforms is the minimum for any real diversification. Size each so that one going to zero still leaves you whole. Most conservative portfolios run 2 to 4 platforms in the 7-9% or 9-11% yield bands, weighted toward the highest-rated names.
Mintos and Capitalia together give you two 80% platforms, both MiFID II or ECSP licensed, both with clean track records. Split 50-50, and one failure costs half your P2P capital. Whether that counts as conservative depends on how much of your total wealth you hold in P2P.
Yes. Higher yields come from platforms with weaker track records or lighter supervision. To reach 11-14% realised without excessive concentration, you typically need 6 platforms, each capped at 20% of the P2P allocation and each independently rated above 60% match.
An 8% target lets you hold just Mintos and Capitalia, both 80% platforms. A 12% target pushes you into names like Indemo, Crowdpear, PeerBerry and Maclear, where the match scores run 60-70%. Holding six of them spreads the elevated risk across multiple entities.
If each platform holds 16-17% of your P2P capital, one failure costs you that slice. Your other five continue. The damage is contained. Compare that with a single-platform portfolio where one failure means 100% loss. Platform spread works only if you actually size the positions so that any one can go to zero.
Recovery timelines vary. Some platforms in workout have returned 30-60% of capital over two to three years. Others have returned nothing. Diversification limits the damage but does not eliminate it.
Mintos scores 80% match and holds a MiFID II licence from Latvijas Banka. Pairing it with another 70%+ platform like Capitalia or InRento gives you two independently supervised entities. Split 50-50, and you have a portfolio where one platform failure loses half. Whether that is safe enough depends on your definition of safe and how much of your total wealth sits in P2P.
Two platforms is the minimum for real diversification. It is not bulletproof, but it is vastly better than holding one.
Loan count handles borrower default. Platform count handles platform failure. For 7-9% realised, hold 2 platforms. For 9-11%, hold 4. For 11-14%, hold 6. Size each so that one going to zero is survivable. Going beyond six platforms adds work without meaningful safety unless you are running a six-figure portfolio. Capital at risk. No compensation scheme covers borrower defaults.