Month 2
Week 8 (19/10/2026)
The First Signs of Portfolio Friction
The fourth week of the Observe Phase was relatively quiet from a portfolio performance perspective. No new allocations were made and no material methodology changes were introduced.
However, several small operational events continued to demonstrate that portfolio outcomes are influenced by more than interest rates and expected returns.
Portfolio Allocation
Benchmark Remains in the Lead
The Benchmark Portfolio remains in first place with a Portfolio Value of €603.84.
Portfolio | Portfolio Value |
Benchmark | €603.84 |
Market Norm | €603.69 |
High Score | €603.43 |
Low Score | €602.91 |
The difference between first and last place remains less than one euro.
Eight weeks into the experiment, portfolio construction differences remain significantly larger than performance differences.
Health Score Divergence Appears
For the first time since the experiment began, the portfolios no longer share identical Health Scores.
The Benchmark Portfolio closed the week with a Health Score of 96%, while all score portfolios remain at 100%.
The change was caused by a single late investment and does not represent a permanent loss. Nevertheless, the event demonstrates how quickly portfolio characteristics can diverge even when performance differences remain small.
This is exactly why the Health Score exists. It attempts to capture changes in portfolio condition before they become visible in Portfolio Value.
Platform Experience
Lande
One of the two investments allocated on 10/09/2026 successfully reached its funding target and became active during the week.
The delayed activation also provided an interesting portfolio observation. When the investment was originally modelled, the individual loan's Estimated XIRR was approximately 12.80%. After adjusting the cash-flow schedule to reflect the actual funding date, the Estimated XIRR declined to approximately 12.22%.
Nothing changed regarding the interest rate, borrower, or contractual terms. The only difference was the time spent in Reserved status before the loan became active.
The second investment remains in Reserved status.
In addition, one investment currently falls within the Late (5-15 days) category.
At this stage, this is simply a portfolio observation and not an indication of permanent loss. The investment remains active and future developments will be documented through the regular reporting process.
The event nevertheless demonstrates how portfolio health characteristics can begin diverging even while portfolio values remain very similar.
EstateGuru, Afranga & Debitum
The platforms continued operating normally during the week. Afranga generated the majority of observed cash flows.
Interest payments were received according to expectations. The previously documented withholding tax mechanism remains visible in the cash flows.
Methodology Updates
No methodology changes were introduced this week.
Version 2.0 remains the active methodology.
Looking Ahead
The first meaningful difference in portfolio health has now appeared, while portfolio values remain remarkably close.
The Observe Phase is gradually shifting attention away from initial portfolio construction and towards factors such as repayment timing, loan status, capital deployment, and portfolio health.
These differences remain small today.
Whether they remain small over the next 52 months is one of the central questions the experiment is designed to explore.
Quiet Week, Useful Observations
The third week of the Observe Phase was relatively uneventful.
No new allocations were made, no repayments were received by the score portfolios, and no methodology changes were introduced.
Even so, several small observations continued to improve understanding of how the participating platforms behave.
Portfolio Allocation
Benchmark Holds the Lead
The Benchmark Portfolio remains in first place for a second consecutive week.
A scheduled €0.41 EG Grow payment increased the Benchmark Portfolio Value to €602.55.
Portfolio | Portfolio Value |
Benchmark | €602.55 |
Market Norm | €601.73 |
High Score | €601.67 |
Low Score | €601.64 |
The spread between first and last place remains less than one euro.
The experiment continues to remind that early portfolio rankings should be interpreted with caution.
Platform Experience
EstateGuru
EG Grow delivered its scheduled monthly payment even though the payment date fell on a Saturday.
This is interesting because several manually selected loans across platforms typically process payments during the following business days when payment dates fall on weekends.
One payment is not enough to draw conclusions, but it will be interesting to observe whether this behaviour remains consistent.
Another observation emerged regarding EstateGuru's AUM fee.
The fee is calculated monthly based on performing outstanding principal rather than interest received. While the fee is relatively small and excluded from the experiment's calculations for practical reasons, investors should remain aware that platform fees can influence realized outcomes over time.
The fee currently appears too small to materially affect the objectives of the experiment, especially given the complexity of allocating it precisely across individual investments.
Lande
The two investments allocated on 10/09/2026, one in the High Score Portfolio and one in the Low Score Portfolio, remain in Reserved status.
At the beginning of the Build Phase, opportunities on Lande often reached their funding targets within hours.
The situation now appears very different. Several opportunities have remained available for extended periods and funding is progressing significantly more slowly.
Regardless of the reason, the observation reinforces one lesson:
Available opportunities and funded opportunities are not always the same thing.
Afranga & Debitum
No payments were expected during the week.
Both platforms continue to operate normally and no material changes were observed.
Methodology Updates
No methodology changes were introduced this week.
Version 2.0 remains the current methodology.
Looking Ahead
Three weeks into the Observe Phase, some of the most interesting observations are not coming from returns but from operational behaviour.
Funding speed, payment timing, fee structures, tax withholding, and capital deployment all influence how an investment performs in practice.
The returns remain small.
The lessons continue to grow.
The Portfolios Start Working
The second week of the Observe Phase produced no major portfolio events, but an important milestone was reached.
For the first time since the experiment began, every score portfolio received interest payments.
Portfolio Allocation
Interest payments were received across all active score portfolios during the week, primarily through Debitum.
Current Leader
The Benchmark Portfolio moves into first place with a Portfolio Value of €602.14.
Portfolio | Portfolio Value |
Benchmark | €602.14 |
Market Norm | €601.73 |
High Score | €601.67 |
Low Score | €601.64 |
The gap between first and last place is just fifty cents.
The race continues to be measured in millimetres rather than metres.
Cash Drag Update
The Benchmark Portfolio continues to benefit from the additional EG Grow allocations completed during the previous week.
Point-in-time Cash Drag now stands at 1%, compared to 13% at the conclusion of the Build Phase.
This significantly reduces one of the largest differences that existed between the Benchmark and the score portfolios.
Platform Experience
Debitum
Debitum became the week's most active platform by providing interest payments across all score portfolios and the Benchmark Portfolio.
While the amounts remain small, they represent the first recurring cash flows generated across all portfolio segments from a single platform.
The payments also highlighted a reporting detail that had previously received little attention.
Several platforms withhold taxes directly at the time interest is paid.
As a result, contractual interest earned and cash available for reinvestment are not always the same amount.
Lande
The two investments allocated on 10/09/2026, one in the High Score Portfolio and one in the Low Score Portfolio, have now remained in Reserved status for more than three weeks.
Methodology Updates
The first week in which all portfolios received taxable interest payments triggered a methodology clarification regarding taxation.
Return-based metrics continue to be reported before personal taxes and withholding taxes in order to improve comparability between platforms and jurisdictions.
Cash-based metrics, however, reflect actual cash available within the portfolio after taxes withheld at source.
Withheld taxes are now reported separately and are excluded from Portfolio Value, Available Cash, and Cash Drag calculations because they are no longer available for reinvestment.
The methodology is now recorded as Version 2.0, while Version 1.0 remains available in the methodology history.
Looking Ahead
The first weeks of the Observe Phase have already demonstrated that portfolio behaviour depends on more than interest rates and expected returns.
Cash Drag, repayment timing, Reserved investments, tax withholding, and platform-specific processes all influence the path between investment selection and realized outcomes.
None of these effects are large on their own.
Combined, they may become far more interesting over time.
The Observe Phase begins quietly
The first week of the Observe Phase is now complete.
After four weeks of portfolio construction, the experiment has entered a very different stage. Capital deployment decisions are largely behind us. The focus now shifts from building portfolios to observing how they behave.
This week was relatively quiet. No new score-based allocations were made, no methodology changes were introduced, and no major portfolio events occurred. Even so, a few operational observations emerged.
Portfolio Allocation
The Benchmark Portfolio was the only portfolio with meaningful investment activity this week.
EstateGuru Grow Reduces Cash Drag
EG Grow allocated an additional €75 during the week through five equal €15 allocations. As a result, Benchmark Cash Drag declined significantly, reaching approximately 8% weighted-average Cash Drag for the week and only 1% point-in-time Cash Drag by week-end.
The additional allocations slightly reduced the portfolio's Weighted Average Interest Rate from 10.97% to 10.59%, reflecting the characteristics of the newly deployed capital.
We Have a New Leader
The High Score Portfolio could not hold the lead for long.
A €0.53 interest payment from EstateGuru pushed the Market Norm Portfolio into first place and made it the first leader of the Observe Phase.
At this stage, the race is measured in cents rather than euros and portfolio differences remain extremely small.
Platform Experience
EstateGuru
The most notable observation this week came from EG Grow.
The platform successfully allocated another €75 of capital through five equal investments. While the process appears to support diversification, it also produced an interesting side effect.
The remaining available balance now stands at €5.01.
According to the published product information, the minimum investment amount per underlying loan is €10. If this remains the case, the remaining balance may stay uninvested until additional capital becomes available. This small amount is unlikely to materially impact performance, but it highlights a practical limitation of automated allocation systems.
Automation can reduce Cash Drag substantially, but it does not always eliminate it completely.
Another observation appeared late in the week.
One of the Low Score Portfolio loans appears to have missed its scheduled repayment date on Sunday. Since the payment date falls on a weekend and the platform has not yet classified the loan as delayed, the situation will simply be monitored during the coming week before any conclusions are drawn.
Lande
Two investments allocated on 10/09/2026, one in the High Score Portfolio and one in the Low Score Portfolio, remain in Reserved status.
This creates an interesting distinction.
Investments remaining in Reserved status for longer periods follow a different lifecycle. Capital is committed to a specific opportunity, but interest only starts accruing once the loan is issued to the borrower. If the funding target is not reached, the investment is cancelled and the reserved capital is returned to the investor.
The capital is no longer available for allocation elsewhere and therefore is not treated as Cash Drag. However, because interest has not yet started accruing, delayed loan issuance may still influence future return metrics such as XIRR.
The lesson is simple:
Committed capital and productive capital are not always the same thing.
Methodology Updates
The methodology remains unchanged and locked.
No material updates were introduced this week. As a result, previously highlighted methodology changes have now been reverted to the default formatting.
Looking Ahead
The first week of the Observe Phase delivered a useful reminder.
Once portfolio construction is complete, small operational details begin to matter. A repayment arriving a few days later than expected, a loan remaining in Reserved status, or a few euros of idle cash can all influence short-term portfolio behaviour.
Nothing significant changed this week.
That is precisely the point.
The portfolios are now built.
The Observe Phase is about watching what happens next.
Month 1
Week 4 (21/09/2026)
Build Phase complete
Four weeks ago, the Score Portfolio Experiment began with a simple objective: build four portfolios with different score characteristics and observe how portfolio construction influences investment outcomes over time.
This week marks the official completion of the Build Phase.
The original plan anticipated approximately €2,000 of invested capital during portfolio construction. The final result exceeded expectations, reaching €2,400 invested capital, creating a larger foundation for the experiment.
More importantly, the methodology has now been finalized and the experiment is ready to transition from portfolio construction to long-term observation.
The experiment will continue with four active platforms:
- EstateGuru
- Afranga
- Lande
- Debitum
SAVY came close to inclusion. Timing had other plans.
Portfolio аllocation
EG Grow allocated two additional €15 positions. Despite this progress, €80.01 remains uninvested. The Benchmark Portfolio entered the week as the only portfolio carrying meaningful Cash Drag.
A more significant event occurred on Lande.
One High Score investment repaid much earlier than expected. The capital was subsequently reinvested using the standard methodology.
Platform | Date | Investment Type | High Score | Nbr of Opportunities | High score | Median Reference Score | Low score | Average Reference Score |
Lande | 16/09/2026 | Re-investment | € 100.27 | 13 | 13.703 | 12.000 | 10.000 | 11.846 |
This also produced the experiment's first Actual Portfolio XIRR. The early repayment on Lande generated an High Score Actual Portfolio XIRR of 5.66%.
The score portfolios enter the Observe Phase with similar characteristics.
The Weighted Average Allocation Scores of the High Score and Market Norm portfolios are now relatively close, while the Low Score Portfolio remains clearly separated.
The Market Norm Portfolio continues to hold the highest Weighted Average Interest Rate, while the High Score Portfolio currently produces the highest Estimated Portfolio XIRR.
The Benchmark Portfolio enters the Observe Phase carrying 13% Cash Drag.
Build Phase baseline:
Portfolio | Portfolio Value | Estimated Portfolio XIRR | Actual Portfolio XIRR | Cash Capital | Active Capital | Portfolio Cash Drag | Portfolio Health Score | Portfolio Default Rate | Portfolio WAVG Allocation Score | Portfolio WAVG Interest % | Portfolio WAVG Duration in Months |
High Score | 600.27 | 12.49% | 5.66% | 0.00 | 600.27 | 0% | 100% | 0.00 | 13.093 | 11.20% | 17.53 |
Market Norm | 600.00 | 12.28% | N/A | 0.00 | 600.00 | 0% | 100% | 0.00 | 12.248 | 11.25% | 24.86 |
Low Score | 600.00 | 9.88% | N/A | 0.00 | 600.00 | 0% | 100% | 0.00 | 10.459 | 9.33% | 20.95 |
Benchmark | 600.01 | N/A | N/A | 80.01 | 520.00 | 13% | 100% | 0.00 | N/A | 10.97% | N/A |
The Build Phase closes with all portfolios healthy, fully operational, and ready to enter long-term observation.
Game of Cents
Last week, the Benchmark Portfolio led by a single cent. This week, the High Score Portfolio moved ahead following the early repayment and reinvestment on Lande. The difference is measured in cents.
The High Score Portfolio enters the Observe Phase as the leader with a Portfolio Value of €600.27. The Benchmark Portfolio follows closely with €600.01.
The difference is insignificant, but the baseline leader is now established.
Platform experience
SAVY
Midway through the week, new opportunities started appearing on SAVY. The timing looked promising. A Paysera account was completed and the final steps of the onboarding process were submitted. Then a message appeared, although all verification and KYC steps had been completed the week before:
Registration completed! We will verify your information within one business day and grant you permission to invest.
Approval did not arrive before the end of Week 4. It also did not arrive on Day 1 of Week 5. At that point, the decision became straightforward. SAVY will not be added to the Score Portfolio Experiment.
The platform remains candidate for the future Proportional Portfolio Experiment.
With this the platform search is now considered complete.
Future platform observations will focus on the four active platforms.
Lande
Lande delivered the most interesting operational lesson of the week. The High Score investment repaid earlier than expected. I did not check each and every day the investments and I noticed the early repayment with a one-day delay. In practice, that means temporary Cash Drag. Lessons learned and now I try to check this occasionally. I was in doubt if Point of Time Cash Drag should be replaced by Weighted average Cash Drag, since it could be influenced by my own portfolio management rather than platform behavior. But that’s part of the experience and methodology remains unchanged.
Two allocations on Lande also remain in Reserved status and have not yet reached their target funding amount. One belongs to the High Score Portfolio and the other belongs to the Low Score Portfolio. Their future funding progress will affect cash flow timing during the Observe Phase. Something I tried to avoid through early allocations during the Build Phase.
EstateGuru
EG Grow continues to leave part of the capital uninvested. As a result, EG Grow remains the primary driver behind the Benchmark Portfolio's 13% Cash Drag. This has become one of the most interesting observations of the Build Phase.
Automation simplified portfolio construction, but did not guarantee immediate capital deployment.
Methodology updates and Website insights
Version 1 of the methodology is now locked. Hopefully, major changes will not be necessary as the experiment progresses.
Here is the Week 4 change overview:
- Platform Exit
The previous Early Exit section has been renamed to Experiment Exit and complemented by a new Platform Exit rule.
Investing should remain rational. Continuing to allocate capital into a deteriorating platform simply to preserve the structure of the experiment would not make sense. A Platform Exit and replacment may therefore be considered when more than 50% of active loans across all portfolios on a single platform are no longer current.
This rule focuses on platform deterioration rather than individual borrower performance.
- Reporting Framework
The conclusion of the Build Phase also clarified what future reporting should look like.
These reporting structures are now documented within the methodology.
Methodology changes introduced during the Week 4 are visually highlighted in blue in the methodology page.
The "Experiment Design" section on the Score Portfolio Experiment home page has been replaced by Key Insights.
As the experiment transitions into the Observe Phase, the home page will place greater emphasis on lessons, observations, and investor takeaways rather than methodology details. The Allocation Score overview will remain the only methodology-related information visible to new visitors, while the full methodology remains available in the dedicated Methodology section.
Looking ahead
The Observe Phase begins now.
The portfolios are built. The methodology is locked. The baseline has been established. Future diary updates will focus less on construction and more on behavior.
The Build Phase was about decisions.
The Observe Phase is about observations. And it begins now!
Week 3 (14/09/2026)
The Build Phase is entering its final week.
Platform аllocation
Each of the four portfolios received an additional €50 per platform, adding €400 to the initial project capital of €2,000.
Platform | Date | Nbr of Opportunities | High score | Median Reference Score | Low score | Average Reference Score |
EstateGuru | 09/09/2026 | 3 | 11.523 | 10.589 | 10.338 | 10.817 |
Debitum | 09/09/2026 | 11 | 15.046 | 14.307 | 12.031 | 13.821 |
Afranga | 10/09/2026 | 4 | 13.703 | 13.007 | 9.966 | 12.421 |
Lande | 10/09/2026 | 7 | 12.000 | 10.592 | 10.000 | 10.852 |
The experiment will continue with four active platforms:
- EstateGuru
- Afranga
- Lande
- Debitum
The original objective was to operate across five platforms.
One lesson became clear during the Build Phase: suitable platforms are significantly harder to find than expected.
Rather than lowering selection standards, the experiment will proceed with a final structure of four platforms and increased allocation sizes. Another reason for this decision is to provide sufficient time for all investments to deploy capital by the end of Week 4 and begin working before the experiment enters the next phase.
Unless a suitable allocation becomes available during the final week of the Build Phase, the four-platform structure is expected to become the final portfolio design.
We have a leader
For the first time, the experiment has a portfolio leader. The lead may only be one cent, but every experiment needs a first leader.
The Benchmark Portfolio currently leads with a Portfolio Value of €600.01. The entire lead comes from the first observed return of €0.01 received through EG Grow. While the difference is negligible, the methodology defines the portfolio with the highest current Portfolio Value as the portfolio leader.
The situation is also somewhat unusual.
The Benchmark Portfolio is currently the only portfolio carrying significant Cash Drag, yet it remains the leader. It is too early for this observation to be meaningful, but it provides the first reminder that portfolio construction and portfolio performance are not necessarily the same thing.
Platform experience
No material changes occurred this week regarding platform selection. Registration restrictions, onboarding requirements, limited opportunity flow, reporting quality, and benchmark compatibility reduced the pool of practical candidates. No reviewed platform provided a sufficiently strong reason to alter the current four-platform structure. I've decided to stop the endless cycle of new registrations and KYC applications.
The platform search is considered substantially complete.
Another observation emerged during analysis of EG Grow. My initial research was that underlying allocations would be transparently visible within the platform, but that is not currently the case.
In fact, EG Grow provides portfolio-level reporting, but detailed information regarding underlying positions, repayment schedules, and diversification is limited compared with manually selected loans. These details currently require a separate request to EstateGuru.
As a result, some Benchmark Portfolio characteristics cannot currently be observed with the same level of precision as the score-selected portfolios.
This appears to be one of the trade-offs between automation and transparency. Even with these reporting limitations, I do not regret choosing EG Grow over the regular Auto Invest option. I still believe it provides a different view of portfolio building. The purpose of the Benchmark Portfolio is to remain largely automated without simply repeating the manual selection process used by the score portfolios.
Methodology updates
Several methodology clarifications were introduced this week.
- Something established at the beginning of the project, but not explicitly stated in the methodology, has now been clarified. The experiment now explicitly states that investments are selected from the primary market. Secondary market purchases are excluded from the methodology.
During the Exit Phase, investments will generally be allowed to mature naturally rather than being actively sold on secondary markets.
- Two additional portfolio characteristics have been added: Number of Unique Investments and Unique Borrowers
The Build Phase has already highlighted that multiple investments do not necessarily represent multiple independent exposures. In some cases, several investments may ultimately be linked to the same borrower. Diversification is not the primary objective of the experiment, but transparency regarding portfolio construction remains important.
These metrics provide additional insight into diversification.
Methodology changes introduced during the Week 3 are visually highlighted in blue in the methodology page.
Looking ahead
The Build Phase started with a simple objective.
Allocate capital according to the methodology.
That objective is now complete.
Week 4 will focus on establishing the final Build Phase baseline before the experiment transitions into the Observe Phase.
The first meaningful observations are beginning to emerge.
Week 2 (07/09/2026)
The Build Phase continued this week
New allocations were completed through Debitum and Lande, bringing total active capital to €1,520 and reducing Cash Drag across all portfolios.
Debitum
Debitum focuses primarily on business financing and sustainable project lending, adding a different type of exposure to the experiment than the platforms participating so far. The platform's reporting differs from the requirements of the Score Portfolio reporting framework. Some metrics and repayment schedules require additional manual reconstruction. Even so, the available information appears sufficient to support the methodology. The Auto Invest strategy is well customizable and aligns well with the project's allocation requirements, which ultimately justified proceeding with the allocation.
Each score portfolio received a new €100 allocation through Debitum following the published methodology.
The Benchmark Portfolio immediately followed a different path. The Debitum Auto Invest configuration used only one rule: "Max. portfolio size" = €100. Debitum automatically distributed capital across ten separate €10 positions, while Lande allocated benchmark capital across two €50 positions.
Platform | Debitum |
Date | 04/09/2026 |
Nbr of Opportunities | 12 |
High score | 15.062 |
Median Reference Score | 14.341 |
Low score | 12.013 |
Average Reference Score | 13.941 |
One observation stood out this week.
The Benchmark Portfolio is becoming significantly more diversified than the score portfolios. While each score portfolio currently holds four investments, the Benchmark Portfolio already holds fifteen. The increased diversification did not result from an intentional diversification target. It emerged naturally from the benchmark products selected by the methodology. At the same time, the Benchmark Portfolio is also carrying more Cash Drag.
This raises an interesting question. Would building and maintaining this level of diversification be practical through manual investment decisions alone? And perhaps more importantly, is greater diversification necessarily better?
Those questions sit at the centre of the Proportional Portfolio Methodology and may become increasingly relevant as both experiments progress.
Platform аllocation
Portfolio values remain equal at €500 and no portfolio leader has emerged.
Portfolio | Portfolio Weighted Average
Allocation Score | Portfolio Weighed Average
Interest % | Portfolio Weighted Average
Duration (Months) |
High Score | 12.953 | 11.00% | 16.44 |
Market Norm | 12.312 | 11.38% | 25.63 |
Low Score | 10.397 | 9.25% | 20.29 |
The first meaningful differences continue to appear in portfolio characteristics rather than portfolio performance.
First Estimated Portfolio XIRR
For the first time, enough investment data is available to calculate Estimated Portfolio XIRR across the score portfolios. Actual Portfolio XIRR is still not available because no investments have completed their lifecycle.
Portfolio | Estimated Portfolio XIRR |
High Score | 12.38% |
Market Norm | 12.47% |
Low Score | 9.74% |
These figures are based on contractual cash flows and assume investments perform according to their published repayment schedules.
The High Score Portfolio now holds a slightly lower weighted average interest rate and XIRR than the Market Norm Portfolio. The difference comes from shorter capital commitment periods, which are rewarded by the Allocation Score formula.
Whether faster capital turnover ultimately outweighs a higher contractual interest rate remains one of the questions I am most curious about.
Platform Experience
Finding suitable platforms continues to be one of the most challenging parts of the Build Phase. EstateGuru, Debitum, Lande, and Afranga are now active participants in the experiment.
The search for a fifth platform continues.
The list of active platforms continues to grow, but so does the list of platforms that looked promising and ultimately failed to fit the methodology. Several candidates were reviewed this week.
Profitus
Profitus was considered, but the platform's automated investment option is currently unavailable. Since passive investment solutions form part of the Benchmark methodology, onboarding has been postponed for now.
Goparity
Goparity remains under consideration. The platform appears compatible with the experiment, but several operational questions remain unresolved.
7Harvests
7Harvests has been added to the watchlist. The available interest rates are substantially higher than most opportunities currently included in the experiment. A small test allocation was made to better understand how the platform behaves. One consideration is that consistently high interest rates could significantly influence the Allocation Score distribution. While the high rates are interesting, I am also looking for greater variation in rates and durations to give the Score Portfolios more room to diverge.
SAVY
SAVY came close to selection. The platform offers extensive Auto Invest customization, but most settings require investor judgment and platform-specific assumptions. Creating a simple and repeatable strategy that could compete fairly with the rest of the portfolios proved difficult, so the platform remains under consideration. A Paysera account is also required, which I do not currently have. For now, the combination of additional setup requirements and methodology design challenges kept SAVY outside the initial selection.
InSoil
InSoil is progressing through onboarding. Three opportunities were available during registration, but account verification was not completed in time. New opportunities will be monitored as they become available. The onboarding process also includes additional KYC requirements. I understand the purpose of these checks, but some of the requested documents would reveal a significant amount of personal financial information not relevant for the purpose of the platform, so I am currently considering my options before proceeding.
Hive5
Hive5 remains under review.
The ideal addition would be another platform offering a broad range of opportunities in customer lending or real estate with transparent reporting and a healthy flow of investment opportunities.
Several well-regarded platforms also remain unavailable due to registration restrictions affecting residents in certain jurisdictions.
The search continues. If you know a platform that may be suitable for the Score Portfolio Experiment, feel free to reach out at contact@donotinitiative.com.
Methodology updates and Website insights
No material methodology changes were introduced this week.
Most of the work focused on improving reporting and presentation.
The Score Portfolio home page has been redesigned to provide a quick overview of the experiment rather than repeating the detailed portfolio comparisons already available on the Dashboard.
Another challenge emerged while calculating Estimated Portfolio XIRR.
Unlike some platforms, Debitum does not provide a month-by-month repayment schedule for each investment. Future cash flows therefore had to be reconstructed manually using the platform's published interest calculation methodology.
The recreated payment schedules currently match the expected repayment amounts shown by Debitum. Even so, the process introduces additional modelling assumptions and will continue to be monitored.
The Benchmark Portfolio presents its own reporting challenges.
Passive products are proving much better at allocating capital than at producing detailed reports.
EstateGuru Grow has already allocated part of the invested capital, but information about the underlying positions remains limited. Hopefully that will change once the funds are transferred to borrowers rather than simply reserved.
With most capital now deployed, the focus is gradually shifting from portfolio construction to portfolio behaviour.
For the first time, the Build Phase feels less like preparation and more like the experiment itself.
Week 1 (31/08/2026)
Finding platforms is harder than building methodologies
The first week of the Build Phase focused on one challenge: deploying capital.
Portfolio allocation and platform experience
The portfolios remain level in terms of Portfolio Value, but their characteristics are already beginning to diverge. The observation period is still too short for many performance metrics to be meaningful.
Portfolio | Portfolio Weighted Average
Allocation Score | Portfolio Weighed Average
Interest % | Portfolio Weighted Average
Duration (Months) |
High Score | 12.25 | 9.83% | 12 |
Market Norm | 11.655 | 11.00% | 28 |
Low Score | 9.858 | 8.67% | 19.33 |
EstateGuru
Week 1 produced the first surprise. The Market Norm allocation selected during Week 0 was cancelled by the platform. Fortunately, another opportunity with the same interest rate and duration was still available from the original opportunity set, so the position was replaced without affecting the methodology.
EG Grow also provided a reminder that capital deployment speed matters. Only €10 out of €100 has been allocated so far.
Platform | EstateGuru |
Date | 23/08/2026 |
Nbr of Opportunities | 8 |
High score | 10.589 |
Median Reference Score | 10.152 |
Low score | 8.041 |
Average Reference Score | 9.918 |
Afranga
Afranga was the most straightforward platform this week. Investment opportunities were available, payment schedules were visible, and both manual and automated investments were processed quickly. Most importantly, the available opportunities created a healthy score spread, which is important for the Score Portfolio methodology.
Platform | Afranga |
Date | 25/08/2026 |
Nbr of Opportunities | 4 |
High score | 13.703 |
Median Reference Score | 13.007 |
Low score | 9.966 |
Average Reference Score | 12.421 |
Lande
Lande required more patience. Opportunities appeared one by one throughout the week and were funded almost immediately. At one point it looked like deployment might not be possible. Then on Friday three loans became available at the same time and provided an opportunity to complete the allocation, so I took it. The score spread was narrower than I had hoped, but waiting for a better setup would have introduced another type of risk: remaining uninvested.
The remaining €100 currently sits in Auto Invest under the default settings and is waiting to be allocated.
Platform | Lande |
Date | 28/08/2026 |
Nbr of Opportunities | 3 |
High score | 12.457 |
Median Reference Score | 12.000 |
Low score | 11.567 |
Average Reference Score | 12.008 |
Several platforms were reviewed but currently do not appear suitable for the experiment.
Mintos
Investing through Notes does not fit well with the reporting structure planned for the experiment. A small test allocation via the Auto-Invest option showed around 7% exposure to loans already 1-15 days late. It wasn't something I was happy to see.
Esketit
Similar to Mintos, the purchase of claim rights introduces additional complexity and makes methodology implementation less straightforward.
Capitalia
The current challenge is a shortage of available investment opportunities.
Still considering other options for a final platform onboarding.
Methodology developments
This week confirmed why the Build Phase exists.
Several refinements were introduced before Version 1.0 becomes locked.
The most important changes were:
- The Build Phase is now formally recognized as a methodology calibration period. Early deployment highlighted several practical constraints that did not appear during methodology design. Formalizing the Build Phase as a calibration period allows these observations to be incorporated before Version 1.0 is locked.
- The experiment now commits a minimum project capital of €2,000 while allowing the final number of participating platforms to depend on platform suitability. The original design assumed a fixed number of platforms. Week 1 showed that platform quality is more important than reaching a predefined target, so the methodology now prioritizes suitability over platform count.
- Weekly Project Diaries will focus on observations and lessons learned. The purpose of the diary is to document the experience of running the experiment as it happens. Operational challenges, surprises, and lessons learned often provide context that performance metrics alone cannot.
- Monthly Reports will focus on data, metrics, and portfolio performance. They will provide a more structured view of portfolio development. Separating observations from performance reporting should make both easier to follow.
- Cash Drag reporting will migrate toward weighted-average measurements during the Observe Phase. A point-in-time Cash Drag figure shows how much capital is uninvested at a specific moment. A weighted-average approach provides a better view of how effectively capital was deployed throughout the reporting period.
Methodology changes introduced during the Week 1 are visually highlighted in blue in the methodology page.
Rethinking transparency
One discussion stood out this week. My initial view was simple:
More data means more transparency.
After reviewing platform feedback and revisiting the purpose of the experiment, I changed my view. The focus should not be to track individual loans. The focus should be to evaluate whether the methodology produces different portfolio outcomes. As a result and if needed, the methodology now allows the use of DONOT-defined identifiers and places greater emphasis on portfolio-level observations. This protects privacy, aligns more closely with platform Terms & Conditins, and keeps attention on the results that matter. Transparency is not about publishing everything. Transparency is about publishing information that helps readers understand the experiment.
Looking ahead
Week 1 reinforced something I had already started noticing in Week 0. Building the methodology was the easy part. Finding suitable places to deploy capital is the harder challenge.
The Build Phase continues.
Will the remaining capital be deployed across four platforms, or will platform availability force further adjustments before Version 1.0 is locked?
Week 2 should provide the first clues.
Week 0 (24/08/2026)
The experiment is live. €2,000 is now at risk. Four portfolios begin with identical capital:
- Low Score Portfolio, €500
- Market Norm Portfolio, €500
- High Score Portfolio, €500
- Benchmark Portfolio, €500
Over the next 60 months, the methodology will decide where that capital goes. Initial deployment has started through EstateGuru. Each portfolio currently holds one €100 position. The remaining capital is waiting for deployment opportunities. The opening allocations are shown below.
Portfolio | Loan ID | Interest % | Duration (Months) | Allocation Score | Amount (€) | Notes |
High Score | #05164356 | 8.5 | 12 | 10.589 | 100 | Highest scoring option |
#85141649 | 9 | 18 | 10.338 | Not selected by any portfolio segment | ||
#72222324 | 9 | 18 | 10.338 | Over Median score. Distance to avg = 0.421. Not selected by Market Norm Portfolio | ||
#15885907 | 9 | 18 | 10.338 | Over Median score. Distance to avg = 0.421. Not selected by Market Norm Portfolio | ||
Market Norm | #39365269 | 8 | 12 | 9.966 | 100 | Under Median Score. Distance to avg = 0.048. Selected using the Tie-Breaker Principle (First Eligible Opportunity) |
#73156093 | 8 | 12 | 9.966 | Under Median Score. Distance to avg = 0.048. Not selected after application of the Tie-Breaker Principle | ||
#27669724 | 8.5 | 18 | 9.764 | Not selected by any portfolio segment | ||
Low Score | #73116790 | 7 | 18 | 8.041 | 100 | Lowest score option |
Benchmark | EG Grow | 7 | 60 | 6.320 | 100 | Passive platform-managed allocation |
Current platform status: Afranga
- Waiting for the incoming bank transfer to be credited.
Lande
- Registration completed. Investor authentication pending. Waiting for additional investment opportunities.
One lesson from Week 0 was unexpected: finding platforms that fit the experiment methodology is proving harder than building the methodology itself.
Do you know a platform that might be suitable for the Score Portfolio Experiment? Send your suggestion to contact@donotinitiative.com.
The Benchmark Portfolio was allocated to EstateGuru Grow using a 60-month commitment period.
EstateGuru Grow allocates capital according to its own investment process, and the duration of the underlying loans is not disclosed at the time of allocation.
As a result, Benchmark Average Allocation Score and Benchmark Average Duration will not be calculated or reported.
The Benchmark Portfolio is intended to serve as a passive investment reference rather than a score-selected portfolio.
All portfolios start level:
- Portfolio Value: €500 • Health Score: 100% • Default Rate: 0% • Portfolio Leader: None
The first week produced an unexpected portfolio selection split before the experiment had even completed its initial deployment.
One question already stands out: Will deployment of the remaining capital strengthen the separation between the score portfolios, or will the portfolios end up looking more similar than expected? The first clues may appear during the Build Phase. Week 1 begins now!