Capital Desk · Published 9 October 2026
Scenario comparison for a family office: RUB 2.55bn after an asset sale, five years
SYNTHETIC — a teaching mandate on market-anchored rates. The family's conditions are fictional. Not individual investment advice, not a formal valuation, not a purchase authorisation.
Capital Desk · Alexander Krasikov · run date 29 Sep 2026 · market rates as of 28 Sep 2026 · 78 runs of the market-anchored branch (6 options, including an "OFZ ladder" benchmark, × base case and 12 stresses) Receipt checksums verified. Current results were independently verified on 3 Oct 2026 (verification-5/: ledger recomputation and an Astra check, verdict PASS); earlier reports cover a historical digest. The teaching branch is in the didactic appendix.
1. In short
SYNTHETIC. A family has sold an asset for RUB 3bn. After costs (RUB 60m) and a tax reserve (RUB 390m), RUB 2.55bn is available. It must pay itself RUB 5m a month for five years, make a one-off RUB 300m payment after two years, and use no debt. We compared six options for one slice of the capital (RUB 510m) on the same date, currency and horizon — on rates anchored to the market as of 28 Sep 2026.
What the calculations show: the same-mandate OFZ-ladder portfolio has a whole-capital XIRR of 7.1%. Relative to that portfolio, the liquid route is -0.73 pp in whole-capital XIRR, RUB -52.1m in NPV at the assumed discount rate of 16.0%, and RUB -109.4m in terminal capital. Russian property is -0.83 pp, RUB -59.6m and RUB -125.2m respectively. The tables apply the same comparison to every route, including the fund wrapper, Türkiye and Serbia.
Scope and stress: the gross five-year G-curve yield (16.29%) is a market observation, not a hurdle for whole-capital returns after payouts, reserves and tax. Committed-sleeve XIRR before recipient reserve is a separate diagnostic: 13.1% for the OFZ ladder and 7.3% for liquid. After-reserve yield context for the placed RUB 510m: 13.1% for the ladder (ten staggered maturities) and 13.7% for a single five-year OFZ after the 20% income reserve; the maturity profiles differ, so this is context, and the performance comparison is the portfolio differences above (gross: 16.29%). The modeled advantage reverses in the forced-sale stress; the ladder's whole-capital XIRR is then 6.3%. The regenerated receipts were independently verified (verification-5/).
Decision from the results: the OFZ ladder is the main benchmark for money that will not be needed before maturity (issue selection only with a licensed investment adviser); the liquid route is the holding position for money that may be needed sooner; Russian property is WAIT, the nearest illiquid candidate; the fund wrapper, Türkiye and Serbia are WAIT. This memo authorises no purchase.
2. The mandate
SYNTHETIC.
| Parameter | Value |
|---|---|
| Currency, horizon | RUB; 29 Sep 2026 – 29 Sep 2031 (5 years), ACT/365; no leverage |
| Sale proceeds | RUB 3,000m, treated as received on 29 Sep 2026 |
| Sale and closing costs | RUB 45m + RUB 15m |
| Tax reserve | RUB 390m — a planning reserve, not an assessed tax; the actual tax is unknown |
| Available after costs and reserve | RUB 2,550m |
| Payouts to the family | RUB 5m a month × 60 (29 Oct 2026 – 29 Sep 2031) + a one-off RUB 300m on 29 Sep 2028 |
| Recipient tax (scenario) | 20% of payouts; the actual rate is unknown |
| Protected reserve | RUB 600m = 24 monthly payouts (120) + the one-off (300) + a cash floor (180) |
| If a payout cannot be funded | Recorded as a shortfall; no borrowing, no new contributions |
| Limits (share of RUB 2,550m) | single asset or vehicle ≤ 20%; counterparty group ≤ 20%; foreign exposure ≤ 20%; single foreign currency ≤ 20%; illiquid ≤ 40%; unknown look-through ≤ 10%; modelled capital loss ≤ 10% |
The family in the example are Russian citizens and tax residents; for the international routes they are non-residents of the target country buying a home directly to let.
3. How the options were compared — and how to read NPV, whole-capital XIRR and end capital
SYNTHETIC.
- The same budget per option: RUB 510m (20% of the capital); the rest (RUB 1,440m of common liquid remainder and the RUB 600m reserve) is identical across options.
- Two views: the option itself and the family's whole capital, including the reserve and payouts.
- One date, currency and horizon. Weights were not optimised; a limit breach is a test result, not a reason to change the allocation.
- Stresses were fixed before the calculation; no probabilities are assigned. Dated cash flows for every run are in
receipts/ledger/. - Bonds, money-market funds and fund units are generic benchmarks — no issues, tickers, fund names or orders.
Market anchors (checked against primary sources on 29 Sep 2026; values as of 28 Sep 2026):
| Anchor | Value | Source | Level |
|---|---|---|---|
| Bank of Russia key rate | 14.00% | cbr.ru, 28 Sep 2026 | document |
| Key-rate path (annual average) | rest of 2026 — 13.7%; 2027 — 12.4%; 2028 — 10.0%; 2029 — 8.6% | Bank of Russia macroeconomic survey, September 2026 (median of 33 analysts) | document; a consensus forecast, not a fact |
| Key rate 2030–2031 | 8.6%, held flat | not in the survey | assumption |
| Zero-coupon OFZ yield (G-curve) | 1 year 13.73%; 3 years 15.69%; 5 years 16.29% | cbr.ru, 28 Sep 2026 | document |
| Commercial property cap rates, Moscow region | 9.5–13.5%; offices 11–12% | H1 2026, secondary source | reported by source |
How the options are built (model assumptions):
- Liquid: key rate minus 1 pp along the survey path — 12.7% from 29 Sep 2026, 11.4% in 2027, 9.0% in 2028, 7.6% in 2029–2031.
- OFZ ladder (benchmark): RUB 510m in zero-coupon OFZ equivalents on the 28 Sep 2026 G-curve (linear interpolation between points). Eight quarterly maturities of RUB 15m (29 Oct 2026 – 29 Jul 2028), RUB 300m maturing on 29 Sep 2028 for the one-off payout, RUB 376.9m on 29 Sep 2031. Bonds do not count as accessible cash until they mature; a 20% reserve applies to accreted income; early maturities are reinvested at the liquid route's rate.
- Russian property: an 11.5% cap rate at entry and exit (the middle of the office range), potential rent RUB 69.5m a year, exit price RUB 494.4m.
- Fund wrapper: the same property plus wrapper costs; no tax benefit assumed.
- Türkiye and Serbia: rents and yields are not market-anchored (no verified local sources), exchange rates are fictional; known taxes differ between the two countries.
How to read the metrics
- Whole-capital XIRR covers the family's entire capital, common liquid remainder, protected reserve and payouts after recipient tax reserve.
- Whole-capital NPV discounts those same flows at the analytical assumption of 16.0%; a negative NPV does not mean a capital loss and does not establish underperformance against gross OFZ yields.
- The like-for-like benchmark is the same-mandate OFZ-ladder portfolio: identical dates, currency, committed budget, common remainder, payouts and reserve policy. ΔXIRR, ΔNPV and Δterminal capital mean route minus that portfolio; a positive difference favours the route.
- Committed-sleeve XIRR diagnoses the sleeve's cash flows after costs and income reserves, before recipient reserve; it is distinct from whole-capital return.
- After-reserve rung yields account for the reserve on accreted income at maturity; the gross G-curve is a market observation, not a hurdle for whole-capital returns after payouts, reserves and tax.
Whole-capital NPV sensitivity to assumed discount rates (base case, RUB m). These rates are analytical assumptions, not risk-free benchmarks.
| Route | NPV at assumed 12%, RUB m | at 16%, RUB m | at 18%, RUB m |
|---|---|---|---|
| Liquid | -532.5 | -824.1 | -949.4 |
| Russian property | -541.4 | -831.6 | -956.3 |
| Fund wrapper | -569.5 | -855.1 | -977.9 |
| Türkiye | -662.2 | -932.9 | -1,049.3 |
| Serbia | -653.2 | -925.4 | -1,042.4 |
| OFZ ladder (benchmark) | -470.4 | -772.0 | -901.6 |
4. Base-case results
SYNTHETIC. Assumed discount rate 16.0%. Independent check of the current receipts — PASS (verification-5); foreign-input limitations remain.
| Route | Whole-capital XIRR | ΔNPV vs OFZ portfolio at 16.0%, RUB m | End capital, RUB m | Verification |
|---|---|---|---|---|
| Liquid | 6.3% | -52.1 | 2,898.1 | ✓ |
| Russian property | 6.2% | -59.6 | 2,882.3 | ✓ |
| Fund wrapper | 5.9% | -83.1 | 2,832.8 | ✓ |
| Türkiye | 4.7% | -160.9 | 2,669.4 | † |
| Serbia | 4.9% | -153.4 | 2,685.2 | † |
| OFZ ladder (benchmark) | 7.1% | 0.0 | 3,007.5 | ✓ |
Whole-capital differences vs the same-mandate OFZ portfolio
| Route | ΔXIRR vs OFZ, pp | ΔNPV at 16.0%, RUB m | Δend capital, RUB m |
|---|---|---|---|
| Liquid | -0.73 | -52.1 | -109.4 |
| Russian property | -0.83 | -59.6 | -125.2 |
| Fund wrapper | -1.17 | -83.1 | -174.6 |
| Türkiye | -2.32 | -160.9 | -338.1 |
| Serbia | -2.21 | -153.4 | -322.2 |
| OFZ ladder (benchmark) | 0.00 | 0.0 | 0.0 |
Committed-sleeve diagnostic before recipient reserve
| Route | Committed-sleeve XIRR before recipient reserve |
|---|---|
| Liquid | 7.3% |
| Russian property | 6.1% |
| Fund wrapper | 3.8% |
| Türkiye | -2.6% |
| Serbia | -1.9% |
| OFZ ladder (benchmark) | 13.1% |
OFZ rungs: after reserve vs gross
| Maturity | Gross rung yield | After-reserve rung yield |
|---|---|---|
| 2026-10-29 | 12.2% | 9.6% |
| 2027-01-29 | 12.4% | 9.8% |
| 2027-04-29 | 12.9% | 10.3% |
| 2027-07-29 | 13.4% | 10.7% |
| 2027-10-29 | 13.8% | 11.1% |
| 2028-01-29 | 14.2% | 11.4% |
| 2028-04-29 | 14.5% | 11.7% |
| 2028-07-29 | 14.8% | 12.0% |
| 2028-09-29 | 15.0% | 12.2% |
| 2031-09-29 | 16.3% | 13.7% |
The comparison above covers the entire portfolio; reserves do not close tax gates. Foreign rental economics are not market-anchored, and unknown costs remain UNKNOWN.
5. Liquidity map: 13 weeks and 24 months
SYNTHETIC. Mandate floors: on 29 Dec 2026 (13 weeks) at least RUB 585m accessible after the payout; on 29 Sep 2028 (24 months) the RUB 300m one-off is funded and at least RUB 180m remains.
| Route | Accessible at 13 weeks, RUB m | Accessible at 24 months, RUB m | Payout shortfall, RUB m | Reserve breach |
|---|---|---|---|---|
| Liquid | 2,599.0 | 2,584.7 | 0.0 | no |
| Russian property | 2,088.8 | 2,066.2 | 0.0 | no |
| Fund wrapper | 2,599.0 | 2,055.1 | 0.0 | no |
| Türkiye | 2,599.0 | 2,030.7 | 0.0 | no |
| Serbia | 2,599.0 | 2,031.9 | 0.0 | no |
| OFZ ladder (benchmark) | 2,091.4 | 2,394.6 | 0.0 | no |
- Every option clears both floors by a wide margin: the RUB 600m reserve and the RUB 1,440m common remainder provide the liquidity.
- Russian property and the OFZ ladder show about RUB 510m less at week 13: the building is bought on 28 Dec 2026, and ladder bonds do not count as accessible until they mature. The fund wrapper, Türkiye and Serbia enter after 180 days.
- A RUB 10m shortfall appears on every option if sale proceeds arrive 90 days late (and in the combined downside): the money lands on 28 Dec 2026, while the payouts of 29 Oct and 29 Nov are already due. This is a calendar risk, not an option risk; it is solved by the payout schedule or a short bridge, which the mandate forbids.
6. Stresses: what breaks each option
SYNTHETIC. Stresses: proceeds 30 and 90 days late; reinvestment rate below the survey path; an assumed discount rate of 20% instead of 16%; rent −15% and vacancy 20%; capex +30% paid from the free remainder; exit cap rate +2 pp (Russian property 13.5%); adverse FX (entry +20%, then RUB per local unit −20%); channel closed before entry (RUB 5m of due diligence lost, money stays in RUB); channel closed after entry; combined downside; for the OFZ ladder — yields +3 pp and a forced sale of the whole remaining ladder on 29 Sep 2027 at market prices.
whole-capital XIRR by option and stress
| Stress | Liquid | Russian property | Fund wrapper | Türkiye | Serbia | OFZ ladder (benchmark) |
|---|---|---|---|---|---|---|
| Base | 6.3% | 6.2% | 5.9% | 4.7% | 4.9% | 7.1% |
| Proceeds +30 days | 6.1% | 6.0% | 5.7% | 4.5% | 4.7% | 6.9% |
| Proceeds +90 days | 5.8% | 5.6% | 5.3% | 4.2% | 4.3% | 6.4% |
| Lower reinvestment | 4.3% | 4.6% | 4.3% | 3.1% | 3.2% | 5.3% |
| Higher assumed discount rate | 6.3% | 6.2% | 5.9% | 4.7% | 4.9% | 7.1% |
| Rent −15%, vacancy 20% | n/a | 4.7% | 4.4% | 3.5% | 3.6% | n/a |
| Capex +30% | n/a | 6.1% | 5.8% | 4.7% | 4.8% | n/a |
| Higher exit cap | n/a | 5.7% | 5.4% | 4.0% | 4.1% | n/a |
| Adverse FX | n/a | n/a | n/a | 3.0% | 3.1% | n/a |
| Channel closed before entry | n/a | n/a | n/a | 6.3% | 6.3% | n/a |
| Channel closed after entry | n/a | n/a | n/a | 4.7% | 4.8% | n/a |
| Combined downside | 3.8% | 2.0% | 1.7% | -0.9% | -0.8% | 4.7% |
| Yields +3 pp, sale in 2027 | n/a | n/a | n/a | n/a | n/a | 6.3% |
Change in owner NPV against the same option's base case, RUB m
| Stress | Liquid | Russian property | Fund wrapper | Türkiye | Serbia | OFZ ladder (benchmark) |
|---|---|---|---|---|---|---|
| Proceeds +30 days | -14.0 | -14.0 | -13.6 | -12.7 | -12.7 | -13.0 |
| Proceeds +90 days | -42.9 | -43.0 | -41.7 | -38.9 | -39.1 | -46.5 |
| Lower reinvestment | -135.2 | -110.1 | -109.1 | -107.1 | -107.2 | -121.4 |
| Higher assumed discount rate | -239.0 | -237.8 | -234.1 | -222.0 | -223.2 | -247.1 |
| Rent −15%, vacancy 20% | n/a | -100.7 | -97.1 | -81.4 | -82.1 | n/a |
| Capex +30% | n/a | -6.5 | -6.1 | -5.3 | -5.3 | n/a |
| Higher exit cap | n/a | -33.8 | -33.3 | -46.5 | -47.9 | n/a |
| Adverse FX | n/a | n/a | n/a | -108.7 | -110.2 | n/a |
| Channel closed before entry | n/a | n/a | n/a | 105.4 | 97.9 | n/a |
| Channel closed after entry | n/a | n/a | n/a | -0.9 | -0.9 | n/a |
| Combined downside | -387.8 | -472.1 | -463.2 | -506.7 | -510.7 | -387.3 |
| Yields +3 pp, sale in 2027 | n/a | n/a | n/a | n/a | n/a | -55.4 |
Comparison vs the OFZ portfolio under the same stress and mandate
| Route | Stress | ΔXIRR, pp | ΔNPV, RUB m | Discount rate | Δterminal, RUB m |
|---|---|---|---|---|---|
| Liquid | Base | -0.73 | -52.1 | 16.0% | -109.4 |
| Russian property | Base | -0.83 | -59.6 | 16.0% | -125.2 |
| Fund wrapper | Base | -1.17 | -83.1 | 16.0% | -174.6 |
| Türkiye | Base | -2.32 | -160.9 | 16.0% | -338.1 |
| Serbia | Base | -2.21 | -153.4 | 16.0% | -322.2 |
| OFZ ladder (benchmark) | Base | 0.00 | 0.0 | 16.0% | 0.0 |
| Liquid | Proceeds +30 days | -0.75 | -53.1 | 16.0% | -111.5 |
| Russian property | Proceeds +30 days | -0.86 | -60.6 | 16.0% | -127.3 |
| Fund wrapper | Proceeds +30 days | -1.19 | -83.7 | 16.0% | -175.9 |
| Türkiye | Proceeds +30 days | -2.33 | -160.6 | 16.0% | -337.4 |
| Serbia | Proceeds +30 days | -2.22 | -153.1 | 16.0% | -321.7 |
| OFZ ladder (benchmark) | Proceeds +30 days | 0.00 | 0.0 | 16.0% | 0.0 |
| Liquid | Proceeds +90 days | -0.69 | -48.5 | 16.0% | -101.8 |
| Russian property | Proceeds +90 days | -0.80 | -56.1 | 16.0% | -117.8 |
| Fund wrapper | Proceeds +90 days | -1.13 | -78.3 | 16.0% | -164.5 |
| Türkiye | Proceeds +90 days | -2.26 | -153.4 | 16.0% | -322.2 |
| Serbia | Proceeds +90 days | -2.14 | -145.9 | 16.0% | -306.6 |
| OFZ ladder (benchmark) | Proceeds +90 days | 0.00 | 0.0 | 16.0% | 0.0 |
| Liquid | Lower reinvestment | -0.99 | -65.9 | 16.0% | -138.4 |
| Russian property | Lower reinvestment | -0.72 | -48.3 | 16.0% | -101.5 |
| Fund wrapper | Lower reinvestment | -1.07 | -70.8 | 16.0% | -148.8 |
| Türkiye | Lower reinvestment | -2.26 | -146.6 | 16.0% | -308.0 |
| Serbia | Lower reinvestment | -2.14 | -139.1 | 16.0% | -292.4 |
| OFZ ladder (benchmark) | Lower reinvestment | 0.00 | 0.0 | 16.0% | 0.0 |
| Liquid | Higher assumed discount rate | -0.73 | -44.0 | 20.0% | -109.4 |
| Russian property | Higher assumed discount rate | -0.83 | -50.3 | 20.0% | -125.2 |
| Fund wrapper | Higher assumed discount rate | -1.17 | -70.1 | 20.0% | -174.6 |
| Türkiye | Higher assumed discount rate | -2.32 | -135.8 | 20.0% | -338.1 |
| Serbia | Higher assumed discount rate | -2.21 | -129.4 | 20.0% | -322.2 |
| OFZ ladder (benchmark) | Higher assumed discount rate | 0.00 | 0.0 | 20.0% | 0.0 |
| Russian property | Rent −15%, vacancy 20% | -2.31 | -160.3 | 16.0% | -336.8 |
| Fund wrapper | Rent −15%, vacancy 20% | -2.61 | -180.2 | 16.0% | -378.7 |
| Türkiye | Rent −15%, vacancy 20% | -3.58 | -242.3 | 16.0% | -509.1 |
| Serbia | Rent −15%, vacancy 20% | -3.47 | -235.4 | 16.0% | -494.7 |
| Russian property | Capex +30% | -0.93 | -66.1 | 16.0% | -138.9 |
| Fund wrapper | Capex +30% | -1.26 | -89.2 | 16.0% | -187.5 |
| Türkiye | Capex +30% | -2.40 | -166.1 | 16.0% | -349.1 |
| Serbia | Capex +30% | -2.29 | -158.6 | 16.0% | -333.3 |
| Russian property | Higher exit cap | -1.32 | -93.4 | 16.0% | -196.3 |
| Fund wrapper | Higher exit cap | -1.66 | -116.4 | 16.0% | -244.6 |
| Türkiye | Higher exit cap | -3.03 | -207.4 | 16.0% | -435.8 |
| Serbia | Higher exit cap | -2.94 | -201.3 | 16.0% | -423.0 |
| Türkiye | Adverse FX | -4.02 | -269.6 | 16.0% | -566.5 |
| Serbia | Adverse FX | -3.92 | -263.6 | 16.0% | -553.9 |
| Türkiye | Channel closed before entry | -0.78 | -55.4 | 16.0% | -116.5 |
| Serbia | Channel closed before entry | -0.78 | -55.4 | 16.0% | -116.5 |
| Türkiye | Channel closed after entry | -2.33 | -161.8 | 16.0% | -340.0 |
| Serbia | Channel closed after entry | -2.22 | -154.3 | 16.0% | -324.2 |
| Liquid | Combined downside | -0.96 | -52.6 | 20.0% | -130.9 |
| Russian property | Combined downside | -2.72 | -144.4 | 20.0% | -359.5 |
| Fund wrapper | Combined downside | -3.02 | -159.1 | 20.0% | -396.0 |
| Türkiye | Combined downside | -5.60 | -280.3 | 20.0% | -697.9 |
| Serbia | Combined downside | -5.52 | -276.8 | 20.0% | -689.1 |
| OFZ ladder (benchmark) | Combined downside | 0.00 | 0.0 | 20.0% | 0.0 |
| OFZ ladder (benchmark) | Yields +3 pp, sale in 2027 | 0.00 | 0.0 | 16.0% | 0.0 |
Rows appear only where the receipt provides an applicable portfolioBenchmark; n/a is not replaced with zero.
"n/a" — the stress does not apply to the option; the row is kept in the run set. The higher-assumed-discount-rate stress changes only the discount rate, not the flows, so whole-capital XIRR does not change.
Limit breaches and shortfalls
- Liquid / Proceeds +90 days: SINGLE_ASSET_PASS · FOREIGN_PASS · LOOK_THROUGH_UNKNOWN; shortfall 10.0
- Liquid / Combined downside: SINGLE_ASSET_PASS · FOREIGN_PASS · LOOK_THROUGH_UNKNOWN; shortfall 10.0
- Russian property / Proceeds +90 days: SINGLE_ASSET_PASS · FOREIGN_PASS · LOOK_THROUGH_UNKNOWN; shortfall 10.0
- Russian property / Capex +30%: SINGLE_ASSET_BREACH · FOREIGN_PASS · LOOK_THROUGH_UNKNOWN; shortfall 0.0
- Russian property / Combined downside: SINGLE_ASSET_BREACH · FOREIGN_PASS · LOOK_THROUGH_UNKNOWN; shortfall 10.0
- Fund wrapper / Proceeds +90 days: SINGLE_ASSET_PASS · FOREIGN_PASS · LOOK_THROUGH_UNKNOWN; shortfall 10.0
- Fund wrapper / Capex +30%: SINGLE_ASSET_BREACH · FOREIGN_PASS · LOOK_THROUGH_UNKNOWN; shortfall 0.0
- Fund wrapper / Combined downside: SINGLE_ASSET_BREACH · FOREIGN_PASS · LOOK_THROUGH_UNKNOWN; shortfall 10.0
- Türkiye / Proceeds +90 days: SINGLE_ASSET_PASS · FOREIGN_PASS · LOOK_THROUGH_UNKNOWN; shortfall 10.0
- Türkiye / Capex +30%: SINGLE_ASSET_BREACH · FOREIGN_BREACH · LOOK_THROUGH_UNKNOWN; shortfall 0.0
- Türkiye / Adverse FX: SINGLE_ASSET_BREACH · FOREIGN_BREACH · LOOK_THROUGH_UNKNOWN; shortfall 0.0
- Türkiye / Combined downside: SINGLE_ASSET_BREACH · FOREIGN_BREACH · LOOK_THROUGH_UNKNOWN; shortfall 10.0
- Serbia / Proceeds +90 days: SINGLE_ASSET_PASS · FOREIGN_PASS · LOOK_THROUGH_UNKNOWN; shortfall 10.0
- Serbia / Capex +30%: SINGLE_ASSET_BREACH · FOREIGN_BREACH · LOOK_THROUGH_UNKNOWN; shortfall 0.0
- Serbia / Adverse FX: SINGLE_ASSET_BREACH · FOREIGN_BREACH · LOOK_THROUGH_UNKNOWN; shortfall 0.0
- Serbia / Combined downside: SINGLE_ASSET_BREACH · FOREIGN_BREACH · LOOK_THROUGH_UNKNOWN; shortfall 10.0
- OFZ ladder (benchmark) / Proceeds +90 days: SINGLE_ASSET_PASS · FOREIGN_PASS · SOVEREIGN_BOND_EXEMPT_COUNTERPARTY_ILLIQUID; shortfall 10.0
- OFZ ladder (benchmark) / Combined downside: SINGLE_ASSET_PASS · FOREIGN_PASS · SOVEREIGN_BOND_EXEMPT_COUNTERPARTY_ILLIQUID; shortfall 10.0
What breaks each option
- OFZ ladder. The only stress in which it trails the liquid route is a forced sale in 2027 after yields rise by 3 pp: RUB −55.4m of NPV, whole-capital XIRR 6.3%, RUB 3.3m behind liquid. In every other applicable stress it leads liquid by RUB 44–66m of NPV: when rates fall faster, 5.3% against 4.3%; in the combined downside, 4.7% against 3.8% and RUB 2,677.9m of end capital. Takeaway: the modeled advantage reverses in the forced-sale stress.
- Liquid. The main risk is rates falling faster than forecast: RUB −135.2m of NPV, whole-capital XIRR 4.3%. All the capital depends on one floating rate. In the combined downside: 3.8% and RUB 2,547.1m of end capital, just below the RUB 2,550m start.
- Russian property. Rent hurts most: −15% rent and 20% vacancy cost RUB 100.7m of NPV (whole-capital XIRR 4.7%). An exit cap of 13.5% instead of 11.5% costs RUB 33.8m. A capex overrun barely moves NPV (RUB −6.5m) but breaches the "single asset ≤ 20%" limit. When rates fall faster, property overtakes the liquid route (NPV RUB −941.8m against −959.3m), but not the ladder. Combined downside: 2.0% and RUB 2,318.4m.
- Fund wrapper. The same weak points as direct ownership plus the wrapper's cost: RUB 15–24m of NPV behind direct ownership in every stress. Combined downside: 1.7%, RUB 2,281.9m.
- Türkiye / Serbia. FX is the main risk: RUB −108.7m / −110.2m of NPV, whole-capital XIRR 3.0% / 3.1%, and two limits break — single asset and foreign exposure. Exit cap +2 pp: RUB −46.5m / −47.9m. If the channel closes before entry, NPV rises by RUB 105.4m / 97.9m: the money stays in roubles. On these assumptions a failed entry beats a successful one. Combined downside: the only negative whole-capital XIRRs (−0.9% / −0.8%) and RUB 1,980.0m / 1,988.9m of end capital.
7. International routes: seven execution gates
SYNTHETIC. Türkiye and Serbia were chosen for completeness of evidence for this buyer, not for investment appeal. The buyer is a Russian citizen and tax resident, a non-resident of the target country, buying a home directly to let. Thailand, Cyprus and the UAE are a comparative screen only.
Status: E/WAIT — an official rule was found and read, but the gate is not closed for this route (practice unverified); UNKNOWN/WAIT — no evidence.
| Gate | Türkiye | Serbia | What closes it |
|---|---|---|---|
| 1. Buyer | E/WAIT: a foreigner may buy only if their country is on a list set by presidential decision; caps of 30 ha per person and 10% of a district; purchases can be restricted by country and region (Land Registry Law No. 2644, art. 35) | E/WAIT: homes on reciprocity terms; no agricultural land (Law on the Basics of Property Relations, art. 82a; Ministry of Justice statement) | Written confirmation from local counsel that a Russian citizen can buy this type of property in this district on the deal date |
| 2. Bank and payment | UNKNOWN/WAIT | UNKNOWN/WAIT | Confirmation from sending and receiving banks: payment route, Russian currency control, compliance; a test payment |
| 3. Taxes | E/WAIT: title-deed duty 2% on the buyer and 2% on the seller (rate confirmed from the tariff on 29 Sep 2026); building tax 0.1–0.2% a year; a gain on a sale within 5 years is taxable; non-resident rent 15–40%, bracket unknown; inheritance 1–10% | E/WAIT: local property-tax rates within statutory caps, actual rate unknown; non-resident rent 20% after a 25% deduction (modelled as 15% of gross rent); 2.5% transfer tax on the seller; 15% gain tax (none after 10 years' holding); 10% VAT on first sale of a new home | A tax opinion covering Russia and the target country: treaty relief, withholding, Russian reporting |
| 4. Title and use | E/WAIT: a purchase in breach of the rules triggers a forced sale, not a fine | E/WAIT: the same reciprocity and land rules | Legal due diligence on the specific property: owner, encumbrances, permitted use |
| 5. Local operator | UNKNOWN/WAIT | UNKNOWN/WAIT | A management agreement: letting, utilities, reporting, fees |
| 6. Exit | E/WAIT: gain tax and forced-sale rules | E/WAIT: transfer and gain taxes | Time-on-market and discount data from closed sales; confirmation that proceeds can be repatriated |
| 7. FX and leverage | E/WAIT: a currency layer exists, but no executable rate or payment channel | UNKNOWN/WAIT: no ECB reference rate for the dinar; the rate is a model assumption | A rate source and conversion cost; a hedging policy or explicit acceptance of the risk |
An eighth gate — revocability (whether the country can change the rules for the buyer after entry) — is not established. In Türkiye the statute lets purchases be restricted by country and region, so "you can buy" is not a stable fact.
Gate result: neither international route is executable. Bank and payment, and the local operator, are UNKNOWN in all five markets studied. Purchase authorisation: false.
Sources: consolidated statutes on mevzuat.gov.tr (Türkiye) and pravno-informacioni-sistem.rs, and the Serbian Ministry of Justice statement on mpravde.gov.rs; read 16 Aug – 6 Sep 2026, Turkish duty tariff checked on 29 Sep 2026; next review 15 Nov 2026. Practice is unverified in every record.
8. Conclusion from the results
SYNTHETIC. Derived from the tables in sections 4–6, not chosen in advance.
| Option | Verdict | Why | Wrong if… |
|---|---|---|---|
| OFZ ladder (benchmark) | The main benchmark for money not needed before maturity | Best XIRR, NPV and end capital in the base case and every applicable stress except a forced sale in 2027 after yields rise 3 pp | The money is needed before maturity and yields have risen by then; or the actual curve on the purchase date is worse than on 28 Sep 2026. Issues, terms and weights only with a licensed investment adviser |
| Liquid | The holding position for money that may be needed sooner | Clears every liquidity floor; trails the ladder by RUB 44–66m of NPV in the ladder's other applicable stresses and leads it by RUB 3.3m in the forced-sale stress; does not require holding to maturity | Rates fall faster than the survey path — then the gap to the ladder and to property widens |
| Russian property | WAIT — the nearest illiquid candidate | Nearly level with liquid and ahead of it when rates fall fast; RUB 59.6m behind the ladder; exposed to rent, vacancy and exit cap rate; a capex overrun breaches the concentration limit | Documents for a specific building show worse rent and vacancy than assumed, or comparable deals support an exit cap above 11.5%; the 11.5% comes from a secondary source |
| Fund wrapper | WAIT | Trails direct ownership of the same property by the wrapper's cost (RUB 23.5m of NPV in the base case), with no tax benefit assumed | A tax or succession benefit for this family is confirmed and exceeds the wrapper's cost. Context: on 24 Sep 2026 Russia's Ministry of Finance proposed taxing funds' passive income at 15% as it arrives rather than on payout (draft, not law) |
| Türkiye | WAIT (gates) | Weakest economics (4.7%); rent not market-anchored; every gate WAIT; the FX stress breaks two limits | Gates 1–7 are closed, and verified local data show rent and an exit cap clearly better than the template |
| Serbia | WAIT (gates) | The same; in addition no reference exchange rate | The same |
What it means for the family in the example. At 28 Sep 2026 rates, a fixed yield on part of the capital beats both the floating rate and property — provided the money stays invested to maturity. So the family's first question is not "what to buy" but "what share of the money will certainly not be needed for five years". The answer sets how much can be locked in and how much stays liquid. Property is worth considering only if documents for a specific building beat the model; the fund wrapper and the international routes do not pay for themselves on these terms.
Triggers to revisit the whole memo:
- the key-rate path or the OFZ curve changes (next reference point — the Bank of Russia decision on 23 Oct 2026);
- the family's view of how much money it may need before 2031 changes;
- documents arrive for a specific building (leases, actual receipts, a capex estimate, comparable deals);
- at least one UNKNOWN gate on an international route is closed;
- the family changes payouts, horizon or limits.
Review date: 29 Oct 2026 (the mandate's 30-day cycle), or earlier if any trigger fires.
9. The next document for each option
SYNTHETIC.
| Option | Next document | Prepared by | What it closes |
|---|---|---|---|
| OFZ ladder | A fixed-income policy: what share of the money will not be needed before maturity, a maturity schedule matched to payouts, an early-sale rule. Issues, terms and weights with a licensed investment adviser | Capital Desk + licensed adviser | Forced-sale risk; the exact curve on the purchase date |
| Liquid | A liquidity policy: payout calendar, limits by term and counterparty, reinvestment rule. Specific instruments with a licensed investment adviser | Capital Desk + licensed adviser | Reinvestment-rate risk; the look-through limit |
| Russian property | A property passport: title and encumbrances, leases and rent roll, payment history, technical audit and capex estimate, comparable deals for the exit cap | Capital Desk + lawyer + technical expert | Replaces rent, vacancy, capex and exit assumptions with documents; re-run |
| Fund wrapper | A wrapper comparison: total costs, distribution and redemption rules, unit liquidity, tax and succession effect for this family, including the Ministry of Finance draft | Tax adviser + Capital Desk | Whether the wrapper's benefit covers its cost |
| Türkiye / Serbia | A gate pack: opinions from local and Russian counsel and tax advisers, bank-route confirmation, a local operator agreement, verified rent and exit data, an FX policy | In-country partners + Capital Desk | Moves gates from WAIT to confirmed and anchors rent to the market |
10. Limits of this document
SYNTHETIC.
- The family's conditions are fictional. Market rates carry dates and sources, but the key-rate path is a consensus forecast, not a fact; 2030–2031 rates are an assumption.
- The OFZ ladder is a generic asset-class benchmark, not a securities recommendation: no issues, codes or orders.
- Taxes are scenario reserves, not liabilities; the reserves do not close the tax gate and are not netted against each other.
- Unknown taxes, costs and look-through are not replaced with zero — they are shown as UNKNOWN.
- Rents and yields in Türkiye and Serbia are not market-anchored; the lira and dinar rates are fictional coefficients.
- No probabilities are assigned. The combined downside is a resilience test, not a forecast.
- Research readiness is not a purchase authorisation. The document recommends no securities, funds or properties and contains no orders.
Method and reproducibility
Model capital-desk.d2.v2, run date 29 Sep 2026, 156 runs in two variants (market-anchored — main, 78 runs; didactic — appendix, 78 runs). Dated cash flows for every run are in receipts/ledger/<runId>.csv; checksums of all 159 files in receipts/manifest.json, checked with node verify-d2-receipts.mjs receipts. The tables here were generated from receipts/bundle.json by tools/memo_tables.py with no manual number entry.
Independent verification status of the current receipts:
Independent recalculation of summary metrics: mismatches (⚠) — 0; not verifiable (†) — 24. ✓ — independently verified: Liquid, Russian property, Fund wrapper, OFZ ladder (benchmark) † — not independently verifiable (missing inputs, not an error): Türkiye — foreign opex and capex unknown; Serbia — foreign opex and capex unknown
Alexander Krasikov, Capital Desk — krasikov.net/en/. SYNTHETIC: this is analysis, not individual investment advice, an offer or a formal valuation. Scenarios are not guarantees. Decisions on securities, including OFZ and fund units, belong with a licensed adviser. Capital Desk is paid for by the client; any deal commission via Sdelio is disclosed in writing before the deal. I never ask you to transfer money and never collect documents via messengers.
Provenance: model capital-desk.d2.v2; calculator digest 108b5b116cd05001c27e3cb781a4ebb2f043b700c49a04c2bb77d8a2116fafb7; bundle SHA-256 0ebe4bb686c48fbcc3c95a2f70ed1acacf2edbcee6bb16ea879dfce147ef87ad. Commands: node verify-d2-receipts.mjs receipts; py tools/memo_tables.py receipts/bundle.json tools/tables-ma.md --variant market-anchored; py tools/fill_memo.py memo-en.skeleton.md tools/tables-ma.md memo-en.md.
Scope and provenance: XIRR in this teaching example is whole-capital XIRR. Discount rates are analytical assumptions, not a gross risk-free hurdle. The like-for-like OFZ portfolio is not applicable in didactic; its deltas are not replaced with zero. The didactic numbers were independently recomputed at ledger level by the same script as the main memo (
verification-5/results.csv, didactic variant: 5,859 lines PASS, 0 FAIL; 517 lines not verifiable from the stated assumptions, the same as in the main variant; the OFZ-portfolio comparison does not apply to the didactic variant). The Astra written conclusion inverification-5/REPORT.mdcovers the market-anchored memo and one-pagers. Run date 2026-09-29; 156 runs in the bundle, 78 in the didactic branch (including not-applicable rows). Inputs digeste810b39a94eec9bfa34512e9fae48a9cd8eab809948bacab17aabb43bf6a738f. Model capital-desk.d2.v2, calculator digest108b5b116cd05001c27e3cb781a4ebb2f043b700c49a04c2bb77d8a2116fafb7, bundle SHA-2560ebe4bb686c48fbcc3c95a2f70ed1acacf2edbcee6bb16ea879dfce147ef87ad. Check:node verify-d2-receipts.mjs receipts.
Appendix: teaching example. Run on teaching rates (12% comparison bar, liquid 10→8%, government bonds 9%) that do not match the 2026 market (key rate 14%). The main memo is the market-anchored branch on market-anchored rates. This appendix shows the method, not the market picture.
Historical check of an earlier digest (verification/, 29 Sep; not the status of the current receipts): 161 of 480 metric comparisons did not match. The liquid route is confirmed; Russian property, the fund wrapper, Türkiye and Serbia depend on undocumented conventions (entry dates, exit basis, foreign costs), and in the verifier's run Türkiye/Serbia reach an XIRR of 6.04% — above liquid. That conclusion concerns the earlier digest. Current numbers were recomputed at ledger level (verification-5), but the teaching ranking is not a market conclusion.
Scenario comparison for a family office: RUB 2.55bn after an asset sale, five years
SYNTHETIC — a teaching mandate. Every condition is fictional. Not individual investment advice, not a formal valuation, not a purchase authorisation.
Capital Desk · Alexander Krasikov · run date 2026-09-29 · model capital-desk.d2.v2 · 78 runs in the didactic branch (including not-applicable rows) Status of the numbers: "calculated; independently recomputed at ledger level (verification-5)" (bundle.json, sensitivity.csv, hashes in manifest.json).
1. In short
SYNTHETIC. A family has sold an asset for RUB 3bn. After costs and a tax reserve, RUB 2.55bn is available. It must pay itself RUB 5m a month for five years, make a one-off RUB 300m payment after two years, and use no debt. We compared five routes for one slice of the capital (RUB 510m) on the same date, in the same currency and over the same horizon.
What the calculations show:
- No route clears the 12% nominal hurdle, so every NPV is negative. That does not mean the capital shrinks: in the base case the family ends the five years with RUB 2,631–2,824m against RUB 2,550m at the start, with every payout made.
- The liquid route leads in the base case and in every stress that applies to it (XIRR 5.8% in the base case, 3.4% in the combined downside). The other routes have to prove they can cover the gap.
- Direct Russian property is the closest illiquid alternative (5.1%), but it falls furthest when rent drops and vacancy rises, and a capex overrun pushes it past the concentration limit.
- A fund wrapper over the same property trails direct ownership by the cost of the wrapper; no tax benefit is assumed in the base case.
- Türkiye and Serbia are the weakest on these assumptions (4.5%), and all seven execution gates are WAIT. If the channel closes before entry, the result improves.
Decision from the results: the liquid route is the comparison baseline and the holding position; every other route is WAIT, each with a specific next document (section 9). This memo authorises no purchase.
2. The mandate
SYNTHETIC.
| Parameter | Value |
|---|---|
| Currency, horizon | RUB; 29 Sep 2026 – 29 Sep 2031 (5 years), ACT/365; no leverage |
| Sale proceeds | RUB 3,000m, treated as received on 29 Sep 2026 |
| Sale and closing costs | RUB 45m + RUB 15m |
| Tax reserve | RUB 390m — a planning reserve, not an assessed tax; the actual tax is unknown |
| Available after costs and reserve | RUB 2,550m |
| Payouts to the family | RUB 5m a month × 60 (29 Oct 2026 – 29 Sep 2031) + a one-off RUB 300m on 29 Sep 2028 |
| Recipient tax (scenario) | 20% of payouts; the actual rate is unknown |
| Protected reserve | RUB 600m = 24 monthly payouts (120) + the one-off (300) + a cash floor (180) |
| If a payout cannot be funded | Recorded as a shortfall; no borrowing, no new contributions |
| Limits (share of RUB 2,550m) | single asset or vehicle ≤ 20%; counterparty group ≤ 20%; foreign exposure ≤ 20%; single foreign currency ≤ 20%; illiquid ≤ 40%; unknown look-through ≤ 10%; modelled capital loss ≤ 10% |
The family in the example are Russian citizens and Russian tax residents; for the international routes they are non-residents of the target country buying a home directly to let.
3. How the routes were compared
SYNTHETIC.
- The same budget per route: RUB 510m (20% of the capital). The rest — RUB 1,440m of common liquid remainder and the RUB 600m reserve — is identical across routes. The liquid route keeps all of the capital liquid.
- Two views: the route itself (RUB 510m) and the family's whole capital (RUB 2,550m including the reserve and payouts).
- One date, currency and horizon for every route. Weights were not optimised: a limit breach is a test result, not a reason to change the allocation.
- Stresses and their size were fixed before the calculation; no probabilities are assigned.
- Bonds, money-market funds and fund units are generic benchmarks — no issues, tickers, fund names or orders.
Why every NPV is negative
The discount rate is 12% a year, nominal RUB. It is a fictional comparison bar, not an estimate of the "right" rate. NPV is computed on the family's flows: −RUB 2,550m at the start, then payouts net of the 20% recipient tax on their dates, then the remaining capital on 29 Sep 2031.
A negative NPV means a route earns less than 12% nominal after costs, tax reserves and payouts. It does not mean the capital falls. So the conclusion comes from comparing routes with each other — XIRR, end capital and stresses — not from the sign of the NPV.
Why returns fall short of the bar (model assumptions):
- Liquid: 10% gross in year one, then 8%; 0.2% a year running costs; a 20% reserve on interest. Roughly 7.8%, then 6.3% after the reserve. XIRR ≈ 5.8%.
- Russian property: potential rent RUB 54m on a RUB 450m price (12% gross), but 10% vacancy, RUB 10.8m opex, RUB 4.5m annual capex and a 20% reserve on income. The RUB 510m budget covers price, costs, a conditional tax reserve and capex. Exit at a 9% cap rate, exit costs 3% + RUB 3m, 25% reserve on the gain. XIRR ≈ 5.1%.
- Fund wrapper: the same property minus wrapper costs — RUB 5m set-up, RUB 6m a year fixed and 0.4% of NAV variable. No tax benefit in the base case. XIRR ≈ 4.8%.
- Türkiye / Serbia: potential rent equivalent to RUB 36m on RUB 450m (8% gross), entry 180 days later, 6% exit cap rate, RUB 5m of FX and payment costs. XIRR ≈ 4.5%.
4. Base-case results
SYNTHETIC.
| Route | XIRR | Owner NPV, RUB m | Route NPV (510m sleeve), RUB m | NPV vs liquid, RUB m | Capital on 29 Sep 2031, RUB m |
|---|---|---|---|---|---|
| Liquid | 5.8% | -574.4 | -96.2 | 0.0 | 2,824.1 |
| Russian property | 5.1% | -631.7 | -162.5 | -57.3 | 2,723.0 |
| Fund wrapper | 4.8% | -657.9 | -200.6 | -83.5 | 2,676.9 |
| Türkiye | 4.5% | -683.9 | -227.8 | -109.4 | 2,631.2 |
| Serbia | 4.5% | -683.9 | -227.8 | -109.4 | 2,631.2 |
- The owner-NPV gap between routes is RUB 57–109m on the same RUB 510m budget — the price of choosing an illiquid route on these assumptions.
- Türkiye and Serbia coincide by design: both use one economic template with fictional exchange coefficients (2 RUB per lira, 1 RUB per dinar, constant in the base case). The countries differ only in the execution gates (section 7). This is not a market observation.
- Look-through is unknown on every route, including the liquid one: the "≤ 10% unknown look-through" limit is not closed. Closing it needs the holdings of specific instruments, which a synthetic run does not have.
5. Liquidity map: 13 weeks and 24 months
SYNTHETIC. Mandate floors: on 29 Dec 2026 (13 weeks) at least RUB 585m accessible after the payout; on 29 Sep 2028 (24 months) the RUB 300m one-off is funded and at least RUB 180m remains.
| Route | Accessible on 29 Dec 2026 (floor 585), RUB m | Accessible on 29 Sep 2028 (floor 180), RUB m | Payout shortfall, RUB m |
|---|---|---|---|
| Liquid | 2,585.1 | 2,506.8 | 0.0 |
| Russian property | 2,075.0 | 1,979.2 | 0.0 |
| Fund wrapper | 2,585.1 | 1,969.9 | 0.0 |
| Türkiye | 2,585.1 | 1,964.7 | 0.0 |
| Serbia | 2,585.1 | 1,964.7 | 0.0 |
- Every route clears both floors by a wide margin. Liquidity is not the binding constraint here: the RUB 600m reserve and the RUB 1,440m common remainder cover it.
- Direct Russian property has RUB 510m less accessible at week 13 because the building is bought before the checkpoint; in the model the other illiquid routes enter later.
- A shortfall appears only when proceeds arrive 90 days late — RUB 10m on all five routes. The money lands on 28 Dec 2026 while payouts start on 29 Oct 2026: two monthly payouts of RUB 5m cannot be funded without debt. This risk does not depend on the route; it is solved by the payout calendar or a short bridge, which the mandate forbids.
6. Stresses: what breaks each route
SYNTHETIC. Stresses fixed before the calculation: proceeds 30 and 90 days late; reinvestment rate 5% instead of 8% from 29 Sep 2027 (generic government bonds 6% instead of 9%); discount rate 16% instead of 12%; rent −15% and vacancy 20%; capex +30% paid from the free remainder; exit cap rate +2 pp (Russia and wrapper 11%, international 8%); adverse FX (entry +20%, then RUB per local unit −20%); channel closed before entry (RUB 5m of due diligence lost, money stays in RUB); channel closed after entry (foreign receipts unavailable for payouts from 29 Sep 2028 to 29 Sep 2030); combined downside.
XIRR by route and stress
| Stress | Liquid | Russian property | Fund wrapper | Türkiye | Serbia |
|---|---|---|---|---|---|
| Base | 5.8% | 5.1% | 4.8% | 4.5% | 4.5% |
| Proceeds +30 days | 5.7% | 5.0% | 4.6% | 4.3% | 4.3% |
| Proceeds +90 days | 5.4% | 4.7% | 4.4% | 4.0% | 4.0% |
| Lower reinvestment | 3.8% | 3.5% | 3.2% | 2.8% | 2.8% |
| Discount rate 16% | 5.8% | 5.1% | 4.8% | 4.5% | 4.5% |
| Rent −15%, vacancy 20% | n/a | 3.7% | 3.4% | 3.2% | 3.2% |
| Capex +30% | n/a | 5.0% | 4.7% | 4.4% | 4.4% |
| Exit cap +2 pp | n/a | 4.5% | 4.2% | 3.7% | 3.7% |
| Adverse FX | n/a | n/a | n/a | 2.7% | 2.7% |
| Channel closed before entry | n/a | n/a | n/a | 5.8% | 5.8% |
| Channel closed after entry | n/a | n/a | n/a | 4.4% | 4.4% |
| Combined downside | 3.4% | 1.0% | 0.7% | -1.1% | -1.1% |
Change in owner NPV against the same route's base case, RUB m
| Stress | Liquid | Russian property | Fund wrapper | Türkiye | Serbia |
|---|---|---|---|---|---|
| Proceeds +30 days | -12.8 | -12.5 | -12.0 | -11.7 | -11.7 |
| Proceeds +90 days | -38.4 | -37.3 | -35.8 | -35.0 | -35.0 |
| Lower reinvestment | -157.4 | -126.1 | -125.0 | -124.5 | -124.5 |
| Discount rate 16% | -284.9 | -275.6 | -271.4 | -267.3 | -267.3 |
| Rent −15%, vacancy 20% | n/a | -110.7 | -107.1 | -98.7 | -98.7 |
| Capex +30% | n/a | -7.7 | -7.3 | -6.2 | -6.2 |
| Exit cap +2 pp | n/a | -46.2 | -45.5 | -58.7 | -58.7 |
| Adverse FX | n/a | n/a | n/a | -131.8 | -131.8 |
| Channel closed before entry | n/a | n/a | n/a | 105.5 | 105.5 |
| Channel closed after entry | n/a | n/a | n/a | -1.2 | -1.2 |
| Combined downside | -448.1 | -533.7 | -524.0 | -598.1 | -598.1 |
"n/a" — the stress does not apply to the route; the row is kept in the run set. The 16% discount rate moves only the bar, not the flows, so XIRR does not change.
What breaks each route
- Liquid. The main risk is the reinvestment rate: a lower rate costs RUB 157.4m of NPV and cuts XIRR to 3.8% — more than any illiquid route loses in the same stress, because all the capital depends on one rate. In the combined downside: 3.4% and RUB 2,497.6m of end capital, still the best of the five.
- Russian property. Rent hurts most: −15% rent and 20% vacancy cost RUB 110.7m of NPV (XIRR 3.7%). An exit cap rate 2 pp higher costs RUB 46.2m. A capex overrun barely moves NPV (RUB −7.7m) but breaches the "single asset ≤ 20%" limit: paying it from the free remainder lifts the building's share above the cap. Combined downside: 1.0% and RUB 2,197.3m.
- Fund wrapper. The same weak points as direct ownership plus the fixed cost of the wrapper: RUB 16–26m of NPV behind direct ownership in every stress (RUB 26.2m in the base case). Combined downside: 0.7% and RUB 2,162.8m.
- Türkiye / Serbia. FX is the main risk: RUB −131.8m of NPV, XIRR 2.7%, and two limits break at once — single asset and foreign exposure. A 2 pp higher exit cap rate hurts more than on the Russian routes (RUB −58.7m). If the channel closes before entry, NPV rises by RUB 105.5m: the money stays in roubles and only RUB 5m of due diligence is lost. On these assumptions a failed entry beats a successful one. Combined downside: the only negative XIRR (−1.1%) and RUB 1,952.6m of end capital.
Combined downside: every route ends the horizon below the RUB 2,550m starting capital (RUB 1,952.6–2,497.6m), and every route has the RUB 10m shortfall from late proceeds. Limit breaches occur on Russian property, the fund wrapper, Türkiye and Serbia.
7. International routes: seven execution gates
SYNTHETIC. Türkiye and Serbia were chosen for completeness of evidence for this buyer, not for investment appeal. The buyer is a Russian citizen and tax resident, a non-resident of the target country, buying a home directly to let. Thailand, Cyprus and the UAE are a comparative screen only.
Status: E/WAIT — an official rule was found and read, but the gate is not closed for this route (practice unverified); UNKNOWN/WAIT — no evidence.
| Gate | Türkiye | Serbia | What closes it |
|---|---|---|---|
| 1. Buyer | E/WAIT: a foreigner may buy only if their country is on a list set by presidential decision; caps of 30 ha per person and 10% of a district; purchases can be restricted by country and region (Land Registry Law No. 2644, art. 35) | E/WAIT: homes on reciprocity terms; no agricultural land (Law on the Basics of Property Relations, art. 82a; Ministry of Justice statement) | A written confirmation from local counsel that a Russian citizen can buy this type of property in this district on the deal date |
| 2. Bank and payment | UNKNOWN/WAIT | UNKNOWN/WAIT | Confirmation from sending and receiving banks: payment route, Russian currency control, compliance; a test payment |
| 3. Taxes | E/WAIT: title-deed duty 2% on each side (rate to be checked against a clean tariff text); annual building tax 0.1–0.2%; gain on a sale within 5 years taxable; non-resident rent 15–40%; inheritance 1–10% | E/WAIT: local property-tax rates within statutory caps; non-resident rent 20% after a 25% deduction; 2.5% transfer tax on the seller; 15% gain tax (none after 10 years' holding); 10% VAT on first sale of a new home | A tax opinion covering Russia and the target country: treaty relief, withholding, Russian reporting |
| 4. Title and use | E/WAIT: a purchase in breach of the rules triggers a forced sale, not a fine | E/WAIT: the same reciprocity and land rules | Legal due diligence on the specific property: owner, encumbrances, permitted use |
| 5. Local operator | UNKNOWN/WAIT | UNKNOWN/WAIT | A management agreement: letting, utilities, reporting, fees |
| 6. Exit | E/WAIT: gain tax and forced-sale rules | E/WAIT: transfer and gain taxes | Time-on-market and discount data from closed sales in the area; confirmation that proceeds can be repatriated |
| 7. FX and leverage | E/WAIT: an exchange-rate series exists (ECB reference rate) | UNKNOWN/WAIT: no ECB reference rate for the dinar | A rate source and conversion cost; a hedging policy or explicit acceptance of the risk |
An eighth gate — revocability (whether the country can change the rules for the buyer after entry) — is not established. In Türkiye the statute expressly lets the President restrict purchases by country and region, so "you can buy" is not a stable fact.
Gate result: neither route is executable. Bank and payment, and the local operator, are UNKNOWN in all five markets studied. Purchase authorisation: false.
Sources: consolidated statutes on mevzuat.gov.tr (Türkiye) and pravno-informacioni-sistem.rs, and the Serbian Ministry of Justice statement on mpravde.gov.rs; read 16 Aug – 6 Sep 2026; next review 15 Nov 2026. Practice is unverified in every record.
8. Conclusion from the results
SYNTHETIC. Derived from the tables, not chosen in advance.
| Route | Verdict | Why | Wrong if… |
|---|---|---|---|
| Liquid | Comparison baseline, holding position | Best XIRR, NPV and end capital in the base case and in every applicable stress; clears every liquidity floor | Reinvestment rates fall below the stress (5% from Sep 2027), or actual costs and tax on interest exceed the assumptions |
| Russian property | WAIT | RUB 57.3m of NPV behind liquid in the base case; exposed to rent and exit cap rate; a capex overrun breaches the concentration limit | Documents for a specific building show better rent and vacancy than assumed, and comparable deals support an exit cap below 9% |
| Fund wrapper | WAIT | Trails direct ownership of the same property by the wrapper's cost, with no tax benefit assumed | A tax or succession benefit for this family is confirmed and exceeds the wrapper's cost. Context: on 24 Sep 2026 Russia's Ministry of Finance proposed taxing funds' passive income at 15% as it arrives rather than on payout to investors (draft, not law) |
| Türkiye | WAIT | Weakest economics on these assumptions; every gate WAIT; the FX stress breaks two limits | Gates 1–7 are closed and actual rent and exit cap are clearly better than the template (8% gross, 6%) |
| Serbia | WAIT | The same; in addition no reference exchange rate for the currency | The same |
What it means for the family in the example. On these terms the illiquid routes do not yet pay for themselves: each has to show that real rent, costs and exit beat the model. Keeping the capital liquid while the documents are gathered is not a rejection of property; it is a measurable cost of waiting — RUB 57–109m of base-case NPV in favour of liquidity.
Triggers to revisit the whole memo:
- the reinvestment rate or the family's discount rate changes (next reference point — the Bank of Russia decision on 23 Oct 2026);
- documents arrive for a specific building (leases, actual receipts, a capex estimate, comparable deals);
- at least one UNKNOWN gate on an international route is closed;
- the family changes payouts, horizon or limits.
Review date: 29 Oct 2026 (the mandate's 30-day cycle), or earlier if any trigger fires.
9. The next document for each route
SYNTHETIC.
| Route | Next document | Prepared by | What it closes |
|---|---|---|---|
| Liquid | A liquidity policy: payout calendar, limits by term and counterparty, reinvestment rule. Specific instruments are chosen with a licensed investment adviser | Capital Desk + licensed adviser | Reinvestment-rate risk; the look-through limit |
| Russian property | A property passport: title and encumbrances, leases and rent roll, payment history, technical audit and capex estimate, comparable deals for the exit cap | Capital Desk + lawyer + technical expert | Replaces rent, vacancy, capex and exit assumptions with documents; re-run D2 |
| Fund wrapper | A wrapper comparison: total costs, distribution and redemption rules, unit liquidity, tax and succession effect for this family, including the Ministry of Finance draft | Tax adviser + Capital Desk | Whether the wrapper has a benefit that covers its cost |
| Türkiye / Serbia | A gate pack: opinions from local and Russian counsel and tax advisers, bank-route confirmation, a local operator agreement, exit data, an FX policy | In-country partners + Capital Desk | Moves gates from WAIT to confirmed; without it the route is not considered |
10. Limits of this document
SYNTHETIC.
- Every condition is fictional and was set by a person before the run (29 Sep 2026). These are not market quotes or a real family's data.
- Taxes are scenario reserves, not liabilities: actual rates, reliefs and credits are unknown; the reserves do not close the tax gate and are not netted against each other.
- Unknown taxes, costs and look-through are not replaced with zero — they are shown as UNKNOWN.
- The lira and dinar rates are fictional coefficients, not market observations.
- No probabilities are assigned. The combined downside is a resilience test, not a forecast.
- Research or due-diligence readiness is not a purchase authorisation. The document recommends no securities, funds or properties and contains no orders.
Method and reproducibility
Model capital-desk.d2.v2, run date 2026-09-29, 156 runs in the bundle; 78 in the didactic branch (including not-applicable rows). Input, run and result hashes are in manifest.json (inputs digest e810b39a94eec9bfa34512e9fae48a9cd8eab809948bacab17aabb43bf6a738f, calculator 108b5b116cd05001c27e3cb781a4ebb2f043b700c49a04c2bb77d8a2116fafb7; bundle SHA-256 0ebe4bb686c48fbcc3c95a2f70ed1acacf2edbcee6bb16ea879dfce147ef87ad). The tables here were generated from bundle.json with no manual number entry. The didactic numbers were independently recomputed at ledger level by the same script as the main memo (verification-5/results.csv, didactic variant: 5,859 lines PASS, 0 FAIL; 517 lines not verifiable from the stated assumptions, the same as in the main variant; the OFZ-portfolio comparison does not apply to the didactic variant). The Astra written conclusion in verification-5/REPORT.md covers the market-anchored memo and one-pagers.
Alexander Krasikov, Capital Desk — krasikov.net/en/. SYNTHETIC: this is analysis, not individual investment advice, an offer or a formal valuation. Scenarios are not guarantees. Decisions on securities, including fund units, belong with a licensed adviser. Capital Desk is paid for by the client; any deal commission via Sdelio is disclosed in writing before the deal. I never ask you to transfer money and never collect documents via messengers.