Asset Tokenization and the Financing Problem Nobody Talks About

I started this research with a simple conviction: Nigeria - and Africa more broadly - has a capital problem that technology might finally be positioned to solve.

Not a shortage of assets. Not a shortage of ambition. Not even a shortage of profitable opportunities. A shortage of the financing architecture needed to turn all three into actual investment.

Walk through any industrial cluster in Lagos, Kano, or Aba and you will find manufacturers sitting on warehouses full of inventory they cannot finance against. Agricultural processors with contractual receivables they cannot discount. Construction firms with land and equipment that banks either undervalue or refuse to touch. These are not failing businesses. They are functioning businesses being systematically starved of the capital they need to grow.

I wanted to understand whether asset tokenization - the idea of converting ownership rights in physical assets into programmable digital instruments - could change that equation. Whether a manufacturer in Apapa with ₦500 million in plant and equipment could one day raise capital against those assets at a price that made expansion rational, without spending six months navigating a banking relationship that might ultimately say no anyway.

That question turned out to be harder to answer than I expected. And the process of trying to answer it rigorously - with 25 years of actual firm-level data from the Nigerian Exchange - taught me things about financing constraints, collateral theory, and the real bottlenecks in Nigerian capital markets that I did not anticipate when I started.

What I found is not a simple vindication of the tokenization thesis. It is something more nuanced, more interesting, and in some ways more useful: evidence about which parts of the tokenization argument hold up against real data, which asset classes show the strongest constraint-reduction signal, and where the regulatory and infrastructure priorities should be concentrated if the goal is genuine economic impact rather than another wave of financial innovation that benefits sophisticated investors while leaving productive firms exactly where they started.

That is what this is about.

The starting point, before any tokenization analysis could begin, was confirming something Nigerian business owners already know but that economics needed to document properly: financing constraints among Nigerian listed firms are severe, pervasive, and structurally embedded.

In economics, the standard way to measure this is called investment-cash flow sensitivity. The logic is clean. In a world where external capital is freely available at a fair price, a firm's investment decisions should depend on how profitable its opportunities are - not on how much cash it happened to generate internally that year. The source of financing should not matter to the investment decision.

But when a firm's investment tracks its internal cash flow almost one-for-one, even after you account for its growth opportunities, that firm is telling you something important through its behaviour: external capital is either unavailable or priced so far above its internal alternative that the firm rationally avoids it. That is a financing constraint. It means investment is not being determined by what projects are worth pursuing - it is being determined by what the firm can self-fund.

Across 105 NGX-listed non-financial firms over 25 years, that sensitivity was statistically significant, economically large, and considerably worse than what firms face in financially developed economies. The coefficient on cash flow in the investment equation - the number that tells you how much a firm increases investment for every additional naira of internal cash - came out at 0.265 after controlling for everything else. In the United States, that number is around 0.22. It sounds close. But the gap compounds across thousands of firms, across decades, across entire sectors of the economy.

Nigerian firms are not just slightly constrained. They are operating with a cost of external capital so high relative to internal funds that the financing source - not the investment opportunity - is driving capital allocation.

That is not an entrepreneurship problem. It is not a management problem. It is a financial architecture problem. And it is the problem that asset tokenization, if designed and implemented correctly, has a genuine theoretical basis for addressing.

Before getting to the data, it is worth being precise about what tokenization means in an economic rather than a technological sense. Because most of the commentary in this space describes it either too narrowly - "putting assets on blockchain" - or too grandly - "democratising finance for everyone." Neither framing is particularly useful for policy or investment decisions.

Tokenization is more accurately described as the transformation of economic rights into programmable, digitally transferable financial claims.

What that means in practice is this: an asset that previously required physical presence, legal intermediaries, bilateral negotiation, and weeks of settlement to transfer or finance against - can instead be represented as a digital instrument that carries its ownership documentation, valuation history, cash flow rights, and encumbrance status with it, is verifiable by any counterparty in real time, and can be transferred or pledged as collateral through automated mechanisms at a fraction of the traditional cost and time.

This changes several things simultaneously. Information asymmetry falls because asset documentation is verifiable rather than trust-based. Transaction costs fall because settlement is automated rather than intermediated. Minimum financing thresholds fall because fractional ownership becomes technically feasible. And liquidity improves because secondary markets can form for assets that were previously too illiquid or too opaque for broad investor participation.

Each of those changes directly addresses one of the specific mechanisms through which financing constraints arise. High information asymmetry makes external equity expensive. High transaction costs make small capital raises uneconomical. Minimum thresholds exclude SMEs. Illiquidity raises the risk premium lenders charge. Tokenization, in theory, attacks all four simultaneously.

That is the economic case. But theory is cheap. The question I wanted to answer was whether you could see any of this in actual firm behaviour - whether firms with assets more structurally suited to tokenization were, in fact, less constrained.

The first methodological challenge was measuring something that does not yet formally exist in Nigeria. Completed tokenization transactions on the NGX are essentially zero. You cannot observe adoption. You cannot track outcomes. What you can observe is asset structure - and from asset structure, you can construct a reasonable approximation of which firms could benefit most if the infrastructure existed.

I built a tokenization eligibility score from three components available in every firm's audited financial statements:

TokenEligibility = 0.5(Fixed Assets/Total Assets) + 0.3(Receivables/Total Assets) + 0.2(Inventory/Total Assets)

The weights reflect the logic of tokenizability, not arbitrary choices.

Fixed assets - property, plant, and equipment - carry the highest weight because they have clear ownership documentation, established professional valuation methods, and the strongest global tokenization precedent. Real estate and equipment tokenization platforms already operate at scale in the United States, European Union, and Singapore. They are the existence proof that this is not a theoretical construct.

Receivables carry a medium weight. They are contractually defined, with specified counterparties and payment schedules. Invoice financing and factoring have existed for centuries. What tokenization adds is programmability - automated settlement through smart contracts, fractional participation, and secondary liquidity for claims that are currently illiquid and bilateral.

Inventory carries the lowest weight because it is the most operationally complex asset to tokenize. Values fluctuate. Quality degrades. Storage creates dependencies. But it is also, as I will come back to, where the most surprising and practically important finding in the study emerged.

The initial results were, frankly, a problem.

The core interaction term - the coefficient that would tell me whether firms with higher tokenization eligibility exhibited lower investment-cash flow sensitivity - came back positive. Not marginally insignificant. Positive. Pointing in the wrong direction.

My first instinct was the data. My second was the model specification. My third instinct, which turned out to be right, was the variables themselves.

When I computed the correlation between tokenization eligibility and asset tangibility - the standard measure used in the collateral literature - I got r = 0.886.

They were essentially the same variable.

Both are dominated by the fixed asset ratio. Fixed assets over total assets is the largest component of tokenization eligibility and is basically what tangibility measures directly. So when both variables were included in the regression alongside their interactions with cash flow, the model had no mathematical basis for distinguishing them. Any constraint-reduction effect the tokenization interaction might have shown was being absorbed by the tangibility term, which was already doing the same econometric work.

This is the kind of problem that looks like a research failure but is actually a substantive finding in disguise.

The r = 0.886 correlation is not just a technical inconvenience. It is making an economic claim about the current state of Nigerian capital markets: in a country without active tokenization infrastructure, the firms that would benefit most from tokenization look exactly like the firms that already benefit from traditional collateral-based financing. Of course they do. The fixed assets that make a firm tokenization-eligible are the same assets that make it creditworthy under conventional banking.

What tokenization promises - theoretically - is to extend those financing benefits to a wider range of firms, at lower cost, with less intermediation friction, and with better secondary liquidity. But in a pre-tokenization environment, the only signal available in balance sheet data is the traditional collateral channel. The two are not yet distinguishable.

The solution was to orthogonalise. I regressed tokenization eligibility on tangibility and extracted the residual - the variation in eligibility that is not explained by conventional physical asset intensity. This residual is driven primarily by the receivables and inventory components of the score, since those are the dimensions where tokenization eligibility and traditional tangibility actually diverge.

What the orthogonalised score captures is the uniquely tokenization-relevant dimension of asset structure: verifiability beyond what lenders already recognise, programmability of cash flow streams, digital transferability, divisibility, and structured liquidity potential. The fraction of asset composition that conventional banking either cannot price efficiently or chooses not to engage with.

When I re-estimated the main model using this orthogonalised score - and particularly when tangibility was excluded as a control variable to prevent re-introducing the collinearity - the interaction coefficient became negative and statistically significant.

Firms with more of the uniquely tokenizable component in their asset structure exhibited lower investment-cash flow sensitivity. Lower revealed financing constraints.

This is not proof that tokenization solves the capital access problem. It is evidence that the specific dimensions of asset structure that tokenization targets - digital verifiability, receivables programmability, inventory financibility - appear in 25 years of Nigerian firm data as genuinely constraint-reducing, once you strip out the general collateral intensity that was obscuring the signal.

That is a meaningfully different and more defensible conclusion than the blanket claim that tokenization will democratise finance. It says: there is something real here, it is located in specific parts of the balance sheet, and it is separable from the traditional collateral effect that we already understood.

Here is the result that surprised me most, and the one I think has the most immediate practical relevance.

When I decomposed the tokenization eligibility score into its three asset class components and tested each one separately, inventory was the only component whose interaction with cash flow was statistically significant.

CF x Inventory Ratio = -0.69, p = 0.015.

Firms with higher inventory ratios exhibit meaningfully lower investment-cash flow sensitivity. The effect size is large - nearly three times the magnitude of the fixed assets interaction, which was not statistically significant on its own.

Think carefully about what this is saying. The asset class I weighted least heavily, the one most commonly overlooked in tokenization discussions dominated by real estate and infrastructure narratives, is the one actually showing up in Nigerian firm data as the constraint-reduction mechanism.

There are several reasons why this might be true, and they reinforce each other.

Inventory is a current asset. It turns over continuously. A firm with high inventory relative to its total assets has a balance sheet that is dynamic - perpetually generating, liquidating, and replenishing a financing base. This creates more financing touch-points than static fixed asset collateral, which you pledge once and service indefinitely. Dynamic assets are, in a real sense, more financeable assets.

More importantly for Nigeria specifically: the inventory financing infrastructure already exists - just not where most people look. AFEX Commodities Exchange, a private-sector operator, has built a functioning warehouse receipt system that enables Nigerian farmers, cooperatives, and traders to store commodities in accredited warehouses and access financing against them. AFEX has raised over $50 million for agri-SMEs through Africa's first warehouse receipt-backed commercial paper - a commodity-backed instrument with 24-hour cash turnaround. The public-sector Nigeria Commodity Exchange has the statutory mandate for this but has yet to achieve full operational capacity after more than two decades. The institutional precedent for inventory tokenization therefore exists through AFEX's proven private-sector infrastructure. What the data may be picking up is not a hypothetical future benefit of tokenization but a constraint-reduction mechanism that already works in practice - and that tokenization would digitalise, scale, and extend to a far broader range of firms than AFEX currently reaches.

This reframes the entire tokenization conversation for Africa.

Most tokenization discussions in emerging markets begin with the glamorous assets. Luxury real estate. Infrastructure bonds. Private equity. Sovereign wealth vehicles. These make sense from a market-sizing perspective - large, illiquid asset classes where fractional ownership creates obvious investor appeal and headline numbers look impressive.

But my data is pointing somewhere different. It is pointing at agricultural inventory in Kano. Manufacturing stock in Lagos. Trade receivables in Aba. Commodity warehouses along the Niger Delta supply chains.

These are not the assets that attract international investor attention. They are the assets where Nigerian firms are actually constrained, where the financing infrastructure already partially exists, and where the incremental cost of improvement is lowest.

Consider the practical difference. Tokenising a ₦10 billion commercial property in Victoria Island requires: a credible digital land registry, a functioning legal framework for digital property rights, professional valuation standards for tokenised fractions, investor protection mechanisms for fractional owners, and secondary market infrastructure to make those fractions liquid. None of that exists yet at the required standard in Nigeria.

Tokenising a warehouse receipt for 500 tonnes of processed cassava starch requires: verification that the inventory exists and is properly stored, a legal link between the digital token and the physical commodity, and a financing counterparty willing to lend against it. Two of those three things - inventory verification systems and commodity-backed lending markets - already work in Nigeria, imperfectly but operationally.

The near-term tokenization opportunity in Nigeria is not the digitalisation of high-value assets for international capital markets. It is the systematisation, standardisation, and scaling of financing mechanisms that already exist for the inventory and receivables-heavy firms that my data identifies as most constrained.

A mid-sized food manufacturer being able to access inventory-backed financing in 72 hours instead of six weeks, at a rate that reflects the actual risk of the underlying commodity rather than the perceived opacity of its balance sheet, using a settlement mechanism that does not require a room full of lawyers - that is a tractable, near-term target. It is not a cryptocurrency story. It is a supply chain finance story with better infrastructure underneath it.

SEC Nigeria's 2024 digital asset rules are a meaningful step. They signal regulatory engagement with the space and provide a legal scaffolding for experimentation. But the rules are necessary, not sufficient.

My data suggests that the constraint-reduction potential of tokenization is concentrated in inventory and receivables - assets where the underlying financing mechanisms already exist but remain expensive, slow, and inaccessible to most firms. The regulatory priority should therefore not be building cryptocurrency exchange infrastructure or enabling speculative token trading. It should be addressing the three prerequisites that make tokenization economically meaningful for productive firms:

Asset verification infrastructure. Tokenization is only as trustworthy as the link between the digital token and the underlying physical asset. Nigeria needs investment in warehouse verification systems, equipment registries, receivables verification platforms, and digital land registries - not because these are blockchain projects but because they are the data infrastructure that makes any form of asset-backed financing more reliable and cheaper. Blockchain is the settlement layer. Verified asset data is the foundation.

Legal clarity on digital collateral. The most important legal question for tokenization in Nigeria is not what a token is - it is whether a creditor can enforce a security interest in a tokenised asset if a borrower defaults. Lenders will not extend credit against tokenised inventory or receivables until that question is answered clearly and tested in Nigerian courts or resolved through specific legislation.

Custody standards. Asset tokenization is fundamentally a trust and verification business. Someone must verify the asset, maintain reserve integrity, confirm legal ownership, and enforce redemption rights. The institutions that develop credible digital custody of real assets - with enforceable redemption, robust reserve management, and integration with existing legal frameworks - will be positioned at the most valuable layer of the tokenization infrastructure stack. Nigerian financial regulators should be developing custody standards for digital asset-backed instruments now, before market development outpaces the framework.

Banks and financial institutions in Nigeria face a choice about how to interpret tokenization. They can treat it as a competitive threat - a mechanism that disintermediates them from lending markets. Or they can treat it as infrastructure that enables them to serve markets they currently cannot serve efficiently.

The second framing is more accurate and more commercially interesting.

The firms my data identifies as most constrained - inventory-heavy manufacturers, agricultural processors, trade-oriented businesses - are not firms that banks are currently serving well. The information asymmetry is too high, the verification costs are too significant, and the collateral is too operationally complex for conventional credit underwriting to work efficiently. These firms either access credit at punishing rates, collateralise non-core assets, or self-finance through retained earnings.

Tokenization infrastructure - specifically, reliable inventory verification systems, programmable receivables documentation, and standardised warehouse receipt frameworks - reduces all three barriers simultaneously. Banks that invest in building or partnering with this infrastructure gain access to a credit market that is currently underserved not because the firms are bad credit risks but because the information costs of assessing them are too high.

The opportunity is not to be replaced by blockchain. It is to use blockchain-adjacent infrastructure to underwrite credit that is currently uneconomical to extend.

I want to be direct about what this research does not prove, because the tokenization space already has more than enough overclaiming.

This study measures tokenization eligibility - the structural potential for tokenization based on observable asset composition - not actual tokenization. The Nigerian Exchange data does not contain a single completed tokenization transaction. Everything here is ex ante evidence about what asset structures look like before any tokenization market exists. The causal mechanism is inferred from observed patterns, not demonstrated through observed outcomes.

The sample covers 105 listed non-financial firms. That is a near-census of the NGX non-financial universe with available data, which gives the findings descriptive authority for that specific population. But listed Nigerian firms are larger, more formal, and more likely to have existing banking relationships than the broader universe of Nigerian enterprises. The firms that would benefit most from tokenization-enabled financing - the unlisted mid-market manufacturers, the growing agricultural processors, the ambitious but informationally opaque SMEs - are probably not in this dataset at all. If anything, the constraint-reduction potential this study identifies is likely understated for the broader economy.

And the collinearity problem, while addressed through orthogonalisation, is real. In a country without active tokenization markets, balance sheet data alone cannot cleanly separate the tokenization channel from the traditional collateral channel. The orthogonalised results are suggestive and directionally consistent. They are not definitive proof.

After 25 years of data, 105 firms, and more regression iterations than I care to count, the conclusion that kept reasserting itself was this:

Asset tokenization is not primarily a trading innovation. It is a collateral quality improvement. It makes existing assets more verifiable, more liquid, more divisible, and more easily integrated into financing mechanisms. The assets do not change. Their accessibility to capital markets changes.

For Nigeria - a country with a large stock of productive but illiquid assets, a bank-dominated financial system that systematically underserves inventory and receivables-heavy firms, a commodities exchange that has already built the institutional precedent for asset-backed financing, and a regulatory framework that is tentatively engaging with digital assets - that framing has real and specific implications.

The firms most constrained are not the ones without assets. They are the ones whose assets the financial system cannot efficiently see, price, or mobilise. Tokenization, if built on the right regulatory and infrastructure foundations, is a mechanism for solving precisely that problem.

Whether it actually gets built that way - whether the regulatory priorities land on asset verification and digital custody rather than on exchange licensing and speculative trading - is a policy choice. It is not determined by the technology.

But the economic foundation for why it should work, grounded in 25 years of actual firm-level data, is stronger than most of the commentary in this space gives it credit for.

That seemed worth documenting carefully.