How Secondary Market Liquidity Really Responds After Auctions
From observation to execution strategy in Uganda’s bond market
Most investors treat auction day as the main event. The data says otherwise.
Across Uganda’s government securities market between September 2025 and March 2026, secondary market turnover on the day after an auction averaged UGX 823 billion — roughly 51% above the baseline. Two days after? UGX 832.5 billion, a 53% premium. On auction day itself, turnover averaged UGX 357.7 billion, 34% below baseline.
The pattern is consistent enough to be useful. But the more important question isn’t whether it exists — it’s why it exists, and what investors should actually do with it.
Auctions are not liquidity events. They are risk distribution events.
This distinction is easy to miss but critical to get right.
When a government auction closes, the market hasn’t produced liquidity — it has allocated risk. Securities have moved from the government’s balance sheet to the primary market. The secondary market, which was essentially paused while investors committed capital to bids, then has to absorb that allocation. That absorption process — settlement, balance sheet adjustment, portfolio rebalancing — is what drives the post-auction trading surge.
In other words, the D+1 and D+2 liquidity spike is a mechanical consequence of how bond markets work, not a coincidence or a sentiment effect. Once you see it that way, the pattern becomes more predictable and more actionable.
What the numbers show
Period Avg turnover vs. baseline Auction day (D0) UGX 357.7 bn −34% D+1 UGX 823.0 bn +51% D+2 UGX 832.5 bn +53%
Total turnover analysed: UGX 69.16 trillion · Coverage: Sep 2025 – Mar 2026
The 1.5× post-auction premium is the average across the full sample, and it persists across multiple auction cycles. This is not a one-off data artefact driven by a single large block trade — the expansion reflects broad participation across the market.
That said, averages in emerging markets can mislead. The appropriate response isn’t to discard the data, but to read it carefully: the direction of the pattern is robust; the magnitude on any individual day will vary with auction quality, macro conditions, and market structure.
The auction cycle, step by step
Every auction follows the same sequence:
Auction → Securities allocated, clearing yield set
Settlement (D+1) → Cash and bonds transfer; banks and asset managers adjust balance sheets
Rebalancing (D+1 to D+2) → Participants redistribute positions — some dealers were over-allocated, some asset managers were underweight; both sides trade to correct this
Secondary market activation → Turnover expands, price discovery improves, bid-offer spreads typically tighten
The key insight here is that liquidity doesn’t appear when allocation happens. It appears when the consequences of allocation start working their way through the system.
Reading liquidity alongside yield
Turnover data alone tells you when to trade. Yield data tells you what you’re buying.
Post-auction, the secondary market is essentially re-testing the clearing yield set at auction. Two scenarios play out differently:
Strong auction (high bid-to-cover, tight yield) + elevated D+1 turnover → confirmation of genuine demand. Yields tend to stabilise or compress marginally as participants build positions with conviction.
Weak auction (low bid-to-cover, wide yield clearing) + elevated D+1 turnover → a warning signal. High turnover here likely reflects dealers offloading excess allocation into a thin book. Yield pressure may continue into D+2 and beyond.
The post-auction window is a genuine execution opportunity in the first scenario. In the second, it can be a trap. The bid-to-cover ratio is the single best early indicator of which regime you’re in.
A three-phase framework
Rather than thinking about individual days, it helps to think about the auction cycle in three distinct phases:
Phase 1 — Pre-auction (positioning) Speculative activity, uneven liquidity, lower conviction. Not the right moment to establish large positions.
Phase 2 — Auction day (allocation) Risk is being distributed. Secondary activity is structurally constrained. Avoid forcing large trades here — you will pay for it in spread.
Phase 3 — D+1 / D+2 (redistribution) Balance sheets adjust, liquidity expands, price discovery improves. This is the primary execution window.
Most investors do the opposite — they focus attention on auction day and then step back. The data suggests they should do almost the reverse.
Turning this into a decision
A simple signal framework puts the analysis to work:
Increase participation when: bid-to-cover is above average, yield is at or below the previous auction level, and D+1 turnover exceeds the baseline.
Maintain base allocation when: the auction was roughly in line with expectations, but no clear signal from secondary markets yet.
Reduce exposure when: bid-to-cover was weak, yield cleared wide, and D+1 turnover is elevated but price is still moving — suggesting redistribution under stress rather than orderly repositioning.
This converts a market pattern into a repeatable decision process. The inputs are publicly observable; the signal updates at each auction cycle.
When the pattern breaks
The D+1/D+2 liquidity premium is conditional, not guaranteed. Four scenarios weaken or eliminate it:
Weak auction demand. When bid-to-cover is low, there simply isn’t enough redistribution capacity to drive meaningful secondary volume. Or if volume is high, it’s distressed selling — not an opportunity.
Large issuance shocks. Supply that overwhelms demand absorption can extend the pricing uncertainty window well past D+2.
Monetary tightening. System-wide liquidity constraints in the banking system reduce the rebalancing activity that drives the pattern.
Concentrated participation. If a small number of players dominate primary allocation, the redistribution dynamic doesn’t materialise. High turnover numbers can mask very thin actual depth.
Monitoring these four conditions at each auction cycle is the risk management layer around the strategy.
The Impala Market view
Uganda’s bond market operates as a two-stage system. Auctions allocate risk; secondary markets determine final ownership. These are separate functions, separated by time, mechanics, and incentive structures.
Liquidity does not peak at the moment of issuance. It peaks when the market begins to absorb and reprice that issuance — typically 24 to 48 hours later.
Investors who treat auctions as execution events will consistently trade in the thinnest part of the cycle. Investors who treat auctions as price discovery signals — and build positions in the secondary market that follows — have a structural edge.
That edge is real, but it’s conditional. It’s strongest when auction demand is credible, secondary participation is broad, and macroeconomic conditions are stable. Weakest when any of those three conditions breaks down. Tracking all three, at every auction cycle, is the minimum standard for using this framework in practice.
Data: Total turnover analysed UGX 69.16 trillion · Average daily turnover UGX 544.6 billion · Coverage Sep 2025 – Mar 2026
Next steps for this analysis: median turnover by tenor · yield-liquidity interaction dataset · real-time signal tracker
Impala Market publishes insights on African Financial Markets for individual investors, investment clubs, and offshore investors.