What Hibernation Did to Savings: The Real Value of Deposits, Bonds and Insurance Contracts Carried Across the Great Ravine by Sleeping and Waking Holders

1. Introduction

The Cape Town Hibernation Trustees' Office held the registry of suspended contracts for holders who entered hibernation before the Great Ravine, and its redemption files, which close with the last waking cohort of the Golden Age recovery, passed to the archives of the UN Planetary Defense Council (PDC) as a single series (Cape Town Hibernation Trustees' Office, 2–97 CE). The Office accepted certificates of deposit, bonds and insurance policies from sleepers on entry, kept them in trust, and presented them for redemption at whichever issuer survived to honour them on the day the sleeper woke, and its rules are known to us only from its own files, so that every description of its practice below is our reconstruction. Alongside the registry the Joint Exchange Clearing Secretariat preserved the clearing and redemption ledgers of the reconstituted exchanges through which the same issuers traded during the recovery (Joint Exchange Clearing Secretariat, 20–97 CE), and an awake holder of an equivalent contract can therefore be found in the same books. We join the two to measure what the Ravine did to the real value of savings carried across it by sleepers and by those who stayed awake.

Real value, meaning the redemption value of a savings contract deflated by the price of a fixed provisioning basket, is used by the PDC Economic Reconstruction Office to size pension and endowment liabilities, by claims offices to settle disputes between waking holders and issuers, and by the desk literature to date the end of the Ravine. It is hard to measure for the Crisis Era because provisioning prices were quoted sparsely and at irregular times, so that a deflator built from the last quote before a date carries a stale price into the calculation (Adeyemi-Strand, 48 DE). Concretely, a sleeper who woke in a week without a recorded quote would be valued at the price of an earlier week, and the error would depend on how fast prices were moving at that moment.

We revisit the transfer account, which holds that the Ravine moved real wealth from sleepers to the waking because sleepers could not reprice, withdraw or reinvest while the issuers they relied on failed and recovered (Lindqvist, 54 DE). We take the opposite side. Sleeping and waking holders of the same contract type who redeemed in the same phase of the recovery lost the same share of real value, and the shortfall that the account relies on comes from what sleepers held and when they woke.

Both archives are now complete in Deterrence Era, Year 57, and the comparison below uses 959 redeemed sleeper contracts and 1,439 redeemed contracts held by awake holders.

2. Sources and Method

The registry lists 1,079 sleeper contracts that we could trace to a named issuer and a certificate value at entry. Of these 959 were redeemed on waking and 120 were lost, meaning that the issuer failed without honouring the certificate or that no claim was ever recorded against it. The Secretariat ledgers give 1,609 traced contracts of the same types held by awake holders at the same issuers, of which 1,439 were redeemed and 170 lost. The three types are deposits, bonds and insurance policies, the last pooling endowment and annuity contracts, as the Office filed them in one series (Sibiya-Falk, 44 DE), and we use the type recorded on the certificate.

A wake cohort is the phase of the recovery in which a contract was redeemed, with boundaries fixed at the two sessions on which the ledgers record the reopening of the full bond list and of the full insurance list, which gives early, middle and late cohorts. We fixed the nine strata formed by type and cohort, and the contrasts described below, before the registry and the ledgers were joined, as we did for the earlier archive (Rautenbach-Zhou, 56 DE), and the code that does so is part of the release.

Each contract's outcome is the natural logarithm of the ratio between its real value at redemption and its real value at entry, where real value is nominal value divided by a provisioning basket index. The index is built from the dated price registers of the Reconstruction Provisioning Board (Reconstruction Provisioning Board, 1–97 CE) and chain-linked between registers in the manner of Adeyemi-Strand (48 DE). Because quotes are asynchronous, the index at any date is a time-weighted interpolation between the quotes that bracket it, and as a sensitivity check we recompute every figure with the index carried forward from the last quote instead.

A raw gap compares the mean outcome of sleeper contracts with that of awake contracts, pooled over all strata, and the adjusted gap is the average of the nine within-stratum differences weighted by the sleeper count in each stratum, so that it states what the sleepers would have lost had they been awake holders of the same types redeemed in the same phase. Stratum variances are combined and multiplied by a design effect of 1.24, which scales standard errors by 1.11, for clustering of contracts within issuers, taken from a mean of 9.0 contracts per issuer and an intra-issuer correlation of 0.03 under the first-order formula (Hartsfield, 49 DE). Every stratum holds at least 68 sleeper and 121 awake contracts, so we use normal critical values and 95% intervals throughout, and heterogeneity across strata is tested with the standard heterogeneity statistic Q for stratum estimates, on eight degrees of freedom.

Under the transfer account the adjusted gap is negative.

3. What the Ravine Did to Every Holder

Awake holders lost real value in every type and in every cohort, and the loss shrank with the date of redemption. Their geometric mean ratio of real value at redemption to real value at entry was 0.56 overall (mean log ratio -0.583), with 0.45 for deposits, 0.56 for bonds and 0.70 for insurance policies, and with 0.39 in the early cohort, 0.55 in the middle cohort and 0.77 in the late cohort.

Sleepers lost more when taken as a whole. Their mean log ratio was -0.709, a geometric mean ratio of 0.49, with 0.39 for deposits, 0.53 for bonds and 0.67 for insurance policies, and with 0.37 in the early cohort, 0.52 in the middle cohort and 0.76 in the late cohort. The raw gap against awake holders is -0.126 (95% CI -0.182 to -0.070, p < 0.001), a real value 12 per cent lower.

The raw gap is a fact about composition. Sleepers held 41.8% of their redeemed contracts as deposits against 34.9% for awake holders, and 40.3% of their contracts were redeemed in the early cohort against 28.3%, as Table 1 shows, and these are the two strata in which the Ravine took the most from both groups.

4. The Gap After Contract Type and Recovery Date

Within strata the sleepers and the waking are indistinguishable. The adjusted gap is -0.016 (95% CI -0.065 to 0.033, p = 0.52), a ratio of real values of 0.98 (95% CI 0.94 to 1.03), and no transfer larger than 6.5 log points, or about 6 per cent of real value, is compatible with the interval (Table 2).

By type the gap is -0.050 for deposits (95% CI -0.129 to 0.029), 0.005 for bonds (95% CI -0.089 to 0.099) and 0.012 for insurance policies (95% CI -0.063 to 0.087). The nine stratum gaps range from -0.110 to 0.070 and do not differ from one another beyond sampling error (Q = 6.70 on 8 degrees of freedom, p = 0.57), and four of the nine are negative.

With the index carried forward from the last quote, the pooled level of sleeper contracts rises from -0.709 to -0.672 and that of awake contracts from -0.583 to -0.538, and the raw gap becomes -0.134, which indicates that prices rose faster around redemption dates than around entry dates, so that stale quotes understated the redemption deflator more than the entry deflator. The adjusted gap moves from -0.016 to -0.015 (95% CI -0.064 to 0.034).

The level of a contract's real value therefore depends on how asynchronous quotes are handled, and the gap between two groups valued on the same dates does not.

5. Contracts That Never Came Back

A transfer could also take place through total loss, if the certificates of sleepers were more often left with failed issuers or went unclaimed. Because lost contracts have no redemption value, the comparison of real values above is conditional on redemption, and the comparison of loss rates is the check that redemption itself did not differ between the groups. Concretely, a sleeper whose issuer failed contributes nothing to Table 2.

Of 1,079 traced sleeper contracts 120 were lost (11.1%), against 170 of 1,609 traced awake contracts (10.6%), a difference of 0.6 percentage points (95% CI -2.1 to 3.2). Within contract type the difference is 0.2 points (95% CI -2.5 to 2.9, p = 0.87), with sleeper loss rates of 9.1%, 16.7% and 6.8% for deposits, bonds and insurance policies against 9.7%, 15.8% and 6.0% for awake holders (Table 3).

Bonds failed most often in both groups.

6. Discussion

The hypothesis that the Great Ravine systematically transferred real wealth from sleepers to the waking is not supported.

Hibernation did not preserve savings either. Sleeping holders redeemed contracts worth 0.49 of their entry value in real terms, a geometric mean over 959 contracts, and awake holders of the same types redeemed in the same phases lost the same share, so that the Ravine and the recovery together left both groups with the same fraction.

The transfer account of Lindqvist (54 DE) compares pooled sleeper and waking books, and the raw gap of -0.126 reproduces its shortfall. What does not survive is the interpretation, because the gap falls to -0.016 when the heavier holding of deposits by sleepers and their earlier redemption are held fixed, and neither the interval nor the test across strata leaves room for a transfer larger than about 6 per cent of real value. We confirm the finding of Sibiya-Falk (44 DE) that insurance policies kept more of their real value than deposits, with ratios of 0.70 and 0.45 for the waking, and we add that the ordering held for sleepers, with 0.67 and 0.39.

We write within the Luo-Ji / Ye framework, whose two axioms fix a civilisation's first need and its pattern of growth and say nothing about how deposits, bonds or policies were repriced, so that the finding above rests on the registry and the ledgers alone and the framework neither predicts it nor forbids it.

The practical consequence is narrow. A claims office that set compensation for waking sleepers on the raw gap would pay out against a shortfall of 12 per cent that comes from what the sleepers held and when they woke, and the adjusted interval excludes any shortfall beyond 6.5 log points that the Ravine inflicted on them in particular.

7. Limitations of the Record

Only contracts that reached the Office are in the registry, so we cannot say how the sleepers differed in wealth from those who stayed awake, and the awake holders are a reweighted comparison group and not individually matched counterparts. Entry values are taken from certificates and not from independent quotes, the nine strata are coarse, and the design effect of 1.24 rests on a first-order formula, so that a larger intra-issuer correlation would widen every interval above. The index is built from Cape Town provisioning quotes and the sleepers' non-contract wealth, such as land and goods, is outside both archives. A contract lost outright has no real value, and for that reason the real-value comparison is conditional on redemption. Hibernators who woke after the registry closed, including the best-known sleepers of the Crisis Era, fall outside the sample, and the result describes only sleepers who woke during the recovery. The design effect treats sleeper and awake contracts at one issuer as independent, whereas a positive correlation across the two groups would partly cancel in the within-stratum difference, so 1.24 is conservative for the gaps.

The release holds the issuer crosswalk that joins registry certificates to ledger entries, the joined panel of 959 and 1,439 redeemed contracts with the 120 and 170 lost ones, the provisioning index under interpolation and under carry-forward, and the scripts behind each table. It sits in the replication archive of the Economic Reconstruction Office and opens to accredited Council researchers. With it a reader can recompute the composition shares of Table 1, the stratum gaps behind the heterogeneity test and the loss rates of Table 3, and can move the cohort boundaries or the stratum weights to see which of the conclusions depend on them. The crosswalk links the Office's redemption files (Cape Town Hibernation Trustees' Office, 2–97 CE) to the recovery ledgers (Joint Exchange Clearing Secretariat, 20–97 CE), and we fixed the strata before the join in the manner of the earlier correlation study (Rautenbach-Zhou, 56 DE).