Worked example
Does tokenization improve the economics of a bond issuance?
We compare the same six annual issuances under today’s process, a modernised conventional process and a proposed tokenized design.
Illustrative caseSynthetic dataNo production results yet
This illustrative case uses synthetic data and follows the same analytical framework available to IEI members — the same engine, the same gates, the same warnings, including the one it trips on purpose. Read the middle column first: it closes most of the gap, and that is the finding most business cases are constructed to avoid.
The answer in 30 seconds
The tokenized design has the stronger modelled operating result.
EUR 900ka year betterMeasured against both today's process and ordinary modernisation.
Most of the improvement is not specific to tokenization — 25% is.
EUR 225ka yearThe rest is process redesign, standardisation and ordinary digitisation, each of which was available without a shared ledger. The wider transformation case may be attractive while the narrower tokenization case is less conclusive.
Nothing here has been measured in production.
The tokenized column is a forecast at its intended operating point. A plan cannot carry a realised claim, however carefully it was built.
The system gains and the institution funding it does not.
EUR 750kshortfall for the funderThe issuer ends up worse off while the consolidated result is positive. A design can be excellent for the market and unfundable for whoever has to build it.
Decision confidence: directional.
Suitable for planning and sensitivity analysis. Not yet sufficient to demonstrate realised performance.
These are findings, not a recommendation. The Lab reports what a case shows and what that supports; the decision belongs to the people who have to live with it.
The operating models
The same initiative costed each way, over one system boundary and one cast of participants — so a difference cannot appear because somebody moved the line.
Current process
EUR 1.15m
annual net cost
How the process runs today, measured or reconstructed from actuals. The starting point, not the benchmark.
Modernised conventional process
EUR 930k
annual net cost
The same outcome with ordinary modernisation and no ledger — an API, a workflow redesign, better data. The column that decides whether tokenization earned its place.
Proposed tokenized design
EUR 250k
annual net cost
The tokenized design at its intended operating point. A plan, and labelled as one.
Lower is better. Figures are annual and rounded.
Production performance: Not yet available. The tokenized design is currently a forecast, not a measured result.
Where does the EUR 900k improvement come from?
Tokenization programmes usually combine several changes at once: process redesign, standardised data, ordinary automation and capabilities specific to the ledger. Separating them stops the whole transformation benefit being credited to tokenization by default.
25% of the modelled improvement is attributed specifically to tokenization. The wider transformation case may be attractive while the narrower tokenization case is less conclusive.
Who creates the value — and who captures it?
A model can improve the economics for the system and still fail commercially, if the institution expected to finance it does not receive enough of the benefit.
The system works; the sponsor's business case does not.
Incentive alignment: misaligned. The principal funder bears more cost than the value it receives, so the model needs a different allocation of fees and benefits before it can be taken to a board.
Compare participant economics across every model
Current process
-€555,000| The issuer | -€945,000 |
| The arranging bank | €285,000 |
| Registrar and paying agent | €105,000 |
| Investors — outside the analysis boundary | -€90,000 |
| Platform vendor — outside the analysis boundary | €0 |
| External counsel — outside the analysis boundary | €260,000 |
Modernised conventional process
-€395,000| The issuer | -€760,000 |
| The arranging bank | €295,000 |
| Registrar and paying agent | €70,000 |
| Investors — outside the analysis boundary | -€90,000 |
| Platform vendor — outside the analysis boundary | €0 |
| External counsel — outside the analysis boundary | €190,000 |
Proposed tokenized design
€40,000| The issuer | -€750,000 |
| The arranging bank | €765,000 |
| Registrar and paying agent | €25,000 |
| Investors — outside the analysis boundary | -€75,000 |
| Platform vendor — outside the analysis boundary | €150,000 |
| External counsel — outside the analysis boundary | €180,000 |
How reliable is this result?
Four different questions, kept apart on purpose. A case can be complete and weakly evidenced, or well evidenced and impossible to compare with anything — and a single score would hide which.
Reconciled historical actuals from before the change, adjusted openly for volume, mix, rates and anything else that moved.
The proposed operating model is estimated, because no production results exist yet.
These numbers describe what happened. They do not establish what caused it.
Suitable for planning and sensitivity analysis. Not yet sufficient to demonstrate realised performance.
Evidence codes and component scores
Completeness and strength measure different things and routinely disagree: a case can answer every required metric with estimates, which is 100 percent complete and weakly evidenced at the same time. The bands above are the governed labels behind the plain-language rows.
What could change the conclusion?
Each of these describes a case that computes perfectly and could still mean something other than it appears to. They are surfaced with the action that would settle them, rather than silently corrected.
- Interest rates
The interest-rate environment differs between the two periods. Liquidity and funding benefits will move with rates whether or not anything else changed.
Action: Rerun the comparison using consistent rate assumptions, and show the liquidity benefit at both.
Everything else worth knowing5 notes
- Current process: No reserve is priced for model uncertainty, so this figure carries its assumptions at full face value.
- Modernised conventional process: No reserve is priced for model uncertainty, so this figure carries its assumptions at full face value.
- Modernised conventional process: Cumulative cash flow never turns positive over the modelled horizon.
- Proposed tokenized design: Cumulative cash flow never turns positive over the modelled horizon.
- Nothing here establishes cause. The honest description is that like was compared with like, not that the ledger produced the difference.
Methodology and calculation details
Every result traces to its inputs, its evidence and the version of the calculation that produced it. Open a group to see how a figure was built.
Operating costs and revenues3 figures
Current process — annual net value before risk-€1,150,001
annualNetValue (§10.1)
| Incremental revenue | €0 |
| Operating cost avoided | €0 |
| Liquidity and funding benefit | €0 |
| Capital benefit | €0 |
| Client benefit actually captured | €0 |
| Recurring operating cost | -€1,150,000 |
| Running the legacy alongside | -€1 |
| Revenue cannibalised | €0 |
TE.REV.TOTAL_INCREMENTALTE.LIQ.FUNDING_COSTTE.CAP.CAPITAL_CHARGETE.COST.RECURRING.TOTALTE.COST.FIXED.DUAL_RUNTE.REV.CANNIBALIZED
Modernised conventional process — annual net value before risk-€930,000
annualNetValue (§10.1)
| Incremental revenue | €0 |
| Operating cost avoided | €0 |
| Liquidity and funding benefit | €0 |
| Capital benefit | €0 |
| Client benefit actually captured | €0 |
| Recurring operating cost | -€890,000 |
| Running the legacy alongside | -€40,000 |
| Revenue cannibalised | €0 |
TE.REV.TOTAL_INCREMENTALTE.LIQ.FUNDING_COSTTE.CAP.CAPITAL_CHARGETE.COST.RECURRING.TOTALTE.COST.FIXED.DUAL_RUNTE.REV.CANNIBALIZED
Proposed tokenized design — annual net value before risk-€340,000
annualNetValue (§10.1)
| Incremental revenue | €0 |
| Operating cost avoided | €0 |
| Liquidity and funding benefit | €510,000 |
| Capital benefit | €0 |
| Client benefit actually captured | €0 |
| Recurring operating cost | -€730,000 |
| Running the legacy alongside | -€120,000 |
| Revenue cannibalised | €0 |
TE.REV.TOTAL_INCREMENTALTE.LIQ.FUNDING_COSTTE.CAP.CAPITAL_CHARGETE.COST.RECURRING.TOTALTE.COST.FIXED.DUAL_RUNTE.REV.CANNIBALIZED
Liquidity and risk4 figures
Current process — risk-adjusted annual value-€1,150,001
riskAdjustedValue (§10.7)
| Annual net value before risk | -€1,150,001 |
| Change in expected loss | €0 |
| Model and uncertainty reserve | €0 |
TE.RISK.EXPECTED_LOSS
- No reserve is priced for model uncertainty, so this figure carries its assumptions at full face value.
Modernised conventional process — risk-adjusted annual value-€930,000
riskAdjustedValue (§10.7)
| Annual net value before risk | -€930,000 |
| Change in expected loss | €0 |
| Model and uncertainty reserve | €0 |
TE.RISK.EXPECTED_LOSS
- No reserve is priced for model uncertainty, so this figure carries its assumptions at full face value.
Proposed tokenized design — risk-adjusted annual value-€250,000
riskAdjustedValue (§10.7)
| Annual net value before risk | -€340,000 |
| Change in expected loss | €150,000 |
| Model and uncertainty reserve | -€60,000 |
TE.RISK.EXPECTED_LOSS
Proposed tokenized design — liquidity and funding benefit€510,000
liquidityBenefit, net of netting given up (§10.6)
| Funded balance released | €950,000 |
| Buffer released | €0 |
| Netting given up | -€380,000 |
| Round-the-clock liquidity cost | -€60,000 |
TE.LIQ.NETTING_CHANGE
Net present value3 figures
Current process — net present value-€4,473,103
npv at 9.0% (§10.2)
| Year 0 | €0 |
| Year 1 | -€1,055,047 |
| Year 2 | -€967,933 |
| Year 3 | -€888,012 |
| Year 4 | -€814,690 |
| Year 5 | -€747,422 |
risk_adjusted:demo-sq
Modernised conventional process — net present value-€4,176,091
npv at 9.0% (§10.2)
| Year 0 | -€900,000 |
| Year 1 | -€511,927 |
| Year 2 | -€782,762 |
| Year 3 | -€718,131 |
| Year 4 | -€658,835 |
| Year 5 | -€604,436 |
risk_adjusted:demo-oc
Proposed tokenized design — net present value-€3,170,725
npv at 9.0% (§10.2)
| Year 0 | -€2,400,000 |
| Year 1 | -€80,275 |
| Year 2 | -€157,815 |
| Year 3 | -€193,046 |
| Year 4 | -€177,106 |
| Year 5 | -€162,483 |
risk_adjusted:demo-tk
Participant and system economics3 figures
Current process — consolidated system value-€555,000
systemValue, reconciled by two routes (§10.11)
| From participant net positions | -€555,000 |
| From resource effects and boundary crossings | -€555,000 |
| Internal transfers eliminated | €505,000 |
c-arrc-regc-lawc-feec-ops-ic-ops-a
Modernised conventional process — consolidated system value-€395,000
systemValue, reconciled by two routes (§10.11)
| From participant net positions | -€395,000 |
| From resource effects and boundary crossings | -€395,000 |
| Internal transfers eliminated | €430,000 |
o-arro-rego-lawo-feeo-ops-io-ops-a
Proposed tokenized design — consolidated system value€40,000
systemValue, reconciled by two routes (§10.11)
| From participant net positions | €40,000 |
| From resource effects and boundary crossings | €40,000 |
| Internal transfers eliminated | €325,000 |
t-arrt-regt-lawt-lict-feet-ops-it-ops-at-liq
Where the improvement comes from1 figure
What the ledger itself caused€225,000
attribute across eight drivers summing to 100 (§11.2)
| Total observed difference | €900,001 |
| Attributable to the programme | €855,001 |
| Attributable to the ledger | €225,000 |
| Unallocated | €0 |
tokenization_nativeprocess_redesignstandardisationdigitizationmarket_rates
Technical audit record
- Calculation version
- 1.0.0
- Taxonomy version
- 1.0.0
- Input hash
- d3e4218c
A snapshot recalculated from its stored inputs under the same versions reproduces this result exactly. That is what lets a figure signed off today still be defended in three years, under the rules that produced it rather than under whatever they become.
Interpretation
What should management conclude?
Everything above is produced by the engine. What follows is a reading of it, written by a person — the Lab reports findings and stops short of recommending, because a tool that has just described its own evidence as directional has no business issuing instructions.
- 1
The proposed tokenized design has the strongest modelled operating result.
It improves the annual economics against both today's process and conventional modernisation. On the modelled figures, this is the better operating model.
- 2
Most of the improvement is not unique to tokenization.
Process redesign, standardised data and ordinary automation account for the majority of the modelled gain. Each of those was available without a shared ledger, and each could be pursued on its own timetable and at lower risk.
- 3
The evidence is not yet operational.
The tokenized column is a forecast, and part of the liquidity benefit follows the rate environment rather than the platform. Both need settling before the number carries a production claim.
- 4
The commercial model is misaligned.
The system creates limited positive value while the institution expected to finance it remains substantially negative. That gap does not close by itself, and it is the most common reason a technically sound design is never funded.
Bottom line: promising, but not ready to scale.
Strengthen the evidence, normalise the external assumptions and redesign the allocation of costs and benefits before approval. None of those is a reason to stop — they are the work between a persuasive model and a fundable one.
Members
Apply the framework to your own initiative
IEI members build private economic cases, compare tokenized and conventional operating models, test the assumptions that move the answer, examine who captures the value, track forecast against actual as a programme runs, and — once there are enough verified results — compare eligible figures with relevant peers.