NSSF Has Distributed Unrealised Gains Before: What History Tells Us About the 22.53% Debate

The debate surrounding NSSF Uganda's record 22.53% interest declaration has increasingly focused on one issue: unrealised investment gains.

NSSF reported approximately UGX 6.51 trillion of total income for FY2025/26. Of this, approximately UGX 3.88 trillion was realised income while approximately UGX 2.62 trillion, or 40.3%, consisted of unrealised fair-value and foreign-exchange gains.

Against that performance, approximately UGX 5.44 trillion was credited to members.

The immediate question is whether a pension fund should convert investment gains that have not yet been realised through a sale into interest that, once credited to members, effectively becomes a permanent member entitlement.

But this is not the first time NSSF has confronted this question.

Looking backwards reveals something important: NSSF has previously incorporated unrealised gains into member returns, has subsequently experienced reversals in those gains, and in more recent years appears to have followed a considerably more conservative relationship between realised income and interest credited.

The debate did not begin with 22.53%

A remarkably similar controversy occurred in FY2017/18.

That year, NSSF declared a then-record 15% interest rate.

Contemporary reporting showed that NSSF's unrealised income had increased dramatically to approximately UGX 560 billion, driven principally by appreciation in listed equities and movements in foreign currencies.

Contemporary reporting also stated explicitly that the return credited to members was determined using both realised and unrealised income.

The Board reportedly recommended 13.5%, not 15%

The 2017/18 episode becomes even more interesting when the process used to determine the interest rate is examined.

Contemporary investigative reporting, citing NSSF Board correspondence and minutes, reported that the Board had recommended an interest rate of 13.5%.

According to that reporting, the Board considered several factors, including member expectations, the Fund's financial position and the need to preserve an appropriate surplus.

The Minister of Finance ultimately declared 15%.

The same reporting indicated that approximately UGX 573 billion of unrealised gains had been included in NSSF's income and that the Board had specifically recognised the importance of unrealised gains arising from currency movements and regional equity markets.

This historical episode is important because it demonstrates that the question being debated today is not merely theoretical.

NSSF policymakers have previously had to decide how much of exceptional investment appreciation should be passed immediately to members and how much should remain within the Fund.

NSSF also had a strong argument for sharing unrealised gains

It would be incomplete to present only the argument against distributing unrealised gains.

There is also a legitimate argument in favour of allowing members to participate in some unrealised investment appreciation.

Consider a simplified example.

Suppose members' savings enable NSSF to purchase shares for UGX 100 billion.

Over many years those shares appreciate to UGX 500 billion, but NSSF continues holding them.

Economically, members' capital has helped generate UGX 400 billion of additional wealth.

Now consider a member who retires before NSSF eventually sells those shares.

If NSSF recognises investment appreciation for accounting purposes but completely excludes it when determining member returns until the asset is sold, that retiring member may leave the Fund without participating fully in economic gains generated during the period in which his or her savings were invested.

NSSF management made essentially this intergenerational-equity argument during the earlier debate.

Then came FY2018/19 — and markets moved the other way

The following year provides an equally important lesson.

NSSF's underlying realised income actually improved.

Realised income increased by approximately 20.4%, from about UGX 1.04 trillion in FY2017/18 to approximately UGX 1.25 trillion in FY2018/19.

But investment markets and currencies moved against the Fund.

NSSF reported an accumulated unrealised loss of approximately UGX 402 billion on foreign-currency and equity positions.

The declared member interest rate fell from 15% to 11%.

This demonstrates the fundamental characteristic of unrealised gains

An unrealised gain can represent genuine economic wealth.

But it can also reverse.

Suppose an investment purchased for UGX 100 billion rises to UGX 160 billion.

NSSF has economically gained UGX 60 billion.

But if the investment subsequently falls to UGX 120 billion before it is sold, UGX 40 billion of that appreciation disappears.

That does not mean the original UGX 60 billion gain was fictitious.

It means that the gain was exposed to market risk until realisation.

The important asymmetry

For a pension fund, this creates an important asymmetry.

This is the central prudential concern.

A reversible increase in the value of an asset can help create a non-reversible increase in member balances.

2022/23 provides an even more dramatic example of reversal risk

The issue appeared again in FY2022/23, although this time from the opposite direction.

NSSF's audited financial statements show foreign-exchange losses of approximately UGX 1.046 trillion.

A major driver was the depreciation of regional currencies against the Uganda shilling, particularly the Kenyan shilling.

The following year those currency movements changed direction sufficiently for NSSF to report substantial foreign-exchange gains.

This is precisely what makes currency revaluation gains and losses different from coupon interest already received on a bond or dividends already paid by a company.

They can be very large, economically meaningful and highly reversible.

What happened when markets recovered in 2023/24?

In FY2023/24, the direction changed again.

NSSF reported approximately UGX 384.9 billion of fair-value gains and approximately UGX 255.1 billion of foreign-exchange gains.

Together, these represented approximately UGX 640 billion of favourable valuation movements.

But something important happened when the member interest declaration is compared with realised income.

NSSF's realised/core income was approximately UGX 2.529 trillion.

Interest credited to members was approximately UGX 2.073 trillion.

The amount credited was therefore still below realised income.

The same broad pattern continued in 2024/25

FY2024/25 also contained substantial valuation movements.

NSSF reported approximately UGX 649 billion of gains on equities.

These were partly offset by a foreign-exchange loss of approximately UGX 274 billion and a loss associated with Umeme of approximately UGX 117 billion.

Nevertheless, realised/core revenue increased to approximately UGX 3.133 trillion.

NSSF credited approximately UGX 2.797 trillion to members at a declared rate of 13.5%.

Once again:

Then FY2025/26 broke the recent pattern

This is where the current debate becomes particularly interesting.

NSSF's realised income increased strongly again, reaching approximately UGX 3.883 trillion.

That was approximately 23.9% higher than the previous year.

By itself, this was an excellent operating and investment result.

But total reported income increased much faster, reaching approximately UGX 6.506 trillion.

The difference was principally explained by approximately UGX 2.623 trillion of unrealised fair-value and foreign-exchange gains.

NSSF subsequently credited approximately UGX 5.444 trillion to members.

The five-year comparison tells the story

And visually, the departure is striking

The UGX 1.56 trillion question

The arithmetic produces perhaps the most important number in the current debate.

Current-year realised/core income alone therefore does not explain the UGX 5.444 trillion credited to members.

That does not automatically mean that exactly UGX 1.561 trillion of current-year unrealised gains was distributed.

The difference could potentially involve a combination of unrealised investment gains, accumulated reserves, prior-period realised surpluses or other adjustments.

A definitive conclusion requires a reconciliation from NSSF showing the precise sources used in determining the amount available for distribution.

But the published figures clearly establish that the FY2025/26 credit materially exceeded current-year realised income.

What happens if next year is not another exceptional year?

Nobody knows how NSSF's investment portfolio will perform next year.

Regional equities may appreciate further. They may remain broadly stable. Currencies may move in NSSF's favour or against it.

It would therefore be inappropriate to assume that the Fund will suffer a market reversal.

But it would be equally inappropriate to assume that the exceptional UGX 2.62 trillion of unrealised gains recorded in FY2025/26 will recur every year.

That is the purpose of prudential financial management: not predicting exactly what markets will do, but ensuring that the Fund remains resilient under less favourable outcomes.

Consider a simple two-year example

The economic risk of the reversal therefore remains somewhere within the Fund.

Ultimately, that risk is borne through future investment returns, reserves, the Fund's balance sheet and the returns available to members who remain invested.

This is fundamentally a question about risk sharing between generations

There are two competing risks.

Neither extreme necessarily produces perfect fairness.

This is where reserves and smoothing become important

The policy choice does not have to be:

"Distribute all unrealised gains"

or

"Never allow members to participate in unrealised gains."

A third approach is to use a transparent reserve and return-smoothing framework.

During exceptional investment years, part of extraordinary appreciation can be retained.

During weaker years, accumulated reserves can help prevent member returns from fluctuating excessively with short-term movements in equity markets and currencies.

Such an approach recognises that a pension fund has a much longer investment horizon than a single financial year.

The historical record makes the reserve question particularly important

NSSF has already experienced substantial valuation movements in both directions:

• large unrealised gains in FY2017/18;
• an approximately UGX 402 billion unrealised adverse movement in FY2018/19;
• approximately UGX 1.046 trillion of foreign-exchange losses in FY2022/23;
• approximately UGX 640 billion of favourable fair-value and foreign-exchange movements in FY2023/24;
• significant equity gains partly offset by foreign-exchange and other losses in FY2024/25; and
• approximately UGX 2.623 trillion of unrealised gains in FY2025/26.

This history demonstrates why unrealised investment performance should not automatically be treated as recurring income.

The issue is not whether NSSF's accounting is false

This distinction remains important.

Accounting standards can require changes in the fair value of investments and foreign-currency positions to be recognised in financial statements.

An investment that genuinely rises in market value has created economic wealth even if it has not yet been sold.

Recognising such a gain is therefore not, by itself, evidence of "creative accounting".

The more difficult question comes afterwards:

What NSSF should disclose

The debate could be substantially clarified through a detailed reconciliation of the 22.53% declaration.

NSSF could show:

The Fund could also disclose its policy for determining how much unrealised appreciation can be distributed, what reserve is retained against future market reversals and how the interests of retiring and continuing members are balanced.

The key question is whether policy has changed

The five-year figures raise a question that deserves a clear answer.

From FY2021/22 through FY2024/25, NSSF credited between approximately 72% and 89% of annual realised/core income to members.

In FY2025/26, the relationship increased abruptly to approximately 140%.

There may be a sound explanation.

NSSF may have accumulated distributable reserves from previous years. It may have deliberately adopted a policy of allowing members to participate more fully in capital appreciation. Or other reserve movements may explain part of the difference.

But without a clear reconciliation, members cannot easily distinguish between those possibilities.

The important question is therefore not merely:

It did report approximately that amount of total income.

The more useful question is:

Conclusion: history supports both sides of the debate

NSSF's history does not support the simplistic conclusion that unrealised gains are fictitious.

Nor does it support the assumption that exceptional unrealised gains can be distributed without considering the possibility of subsequent reversal.

The FY2017/18 experience demonstrates the case for allowing members to participate in genuine capital appreciation.

The subsequent FY2018/19 reversal, and the much larger foreign-exchange losses experienced in FY2022/23, demonstrate why valuation gains should not be assumed to move permanently in one direction.

The recent five-year numbers make FY2025/26 particularly noteworthy.

In each of the previous four years, interest credited remained below realised/core income.

In FY2025/26, approximately UGX 5.444 trillion was credited against approximately UGX 3.883 trillion of realised income.

That is equivalent to approximately 140.2% of realised income.

At the same time, approximately UGX 2.623 trillion of NSSF's reported income consisted of unrealised gains.

This does not prove that the 22.53% declaration was imprudent.

But it does make the Fund's reserve policy, smoothing policy and treatment of unrealised gains central to understanding the sustainability of the declaration.

NSSF cannot know today whether its investments will appreciate or depreciate next year.

That uncertainty is precisely the point.

A long-term pension fund does not need to predict the next market movement. It needs a policy capable of protecting different generations of members whether markets rise or fall.

The most useful response to the current controversy would therefore be greater disclosure: show members exactly how the UGX 5.444 trillion was funded, how much unrealised appreciation was incorporated, how much was drawn from accumulated reserves, and what buffer remains if some of the extraordinary FY2025/26 gains subsequently reverse.

That would move the debate away from the emotionally attractive question of whether 22.53% is "too high" and towards the more important question: whether NSSF has a transparent and sustainable framework for sharing investment gains fairly between today's members and tomorrow's.

Sources and Further Reading

NSSF Uganda — FY2025/26 Integrated Report
Five-year performance review and financial and operational highlights, including realised income, total income, unrealised gains and interest credited to members.

NSSF Uganda — FY2024/25 Annual Integrated Report
Information on realised revenue, equity gains, foreign-exchange movements and the 13.5% member return.

NSSF Uganda — FY2023/24 Annual Integrated Report and Audited Financial Statements
Information on fair-value gains and foreign-exchange gains and losses, including the substantial FY2022/23 foreign-exchange loss.

The Independent Uganda — "NSSF Courts Savers"
Contemporary reporting on the FY2017/18 declaration, unrealised income and the incorporation of realised and unrealised income in the member return.

Contemporary reporting on the FY2017/18 NSSF interest declaration
Reporting on the Board's recommended rate, the eventual 15% declaration and the debate surrounding unrealised investment gains.

Contemporary reporting on NSSF FY2018/19 performance
Reporting on the increase in realised income alongside approximately UGX 402 billion of unrealised losses on equity and foreign-currency positions.