The tenor was the problem, not the ladder

A monthly saving is a small decision made often, which is exactly why the discipline matters. The Ghost does not ask what is hot. It asks three duller questions, in order.

4.0m to Government T-bonds. Fixed income is the sleeve furthest below its band middle, and the official Bank of Tanzania curve of 28 August is the reason it is worth filling rather than merely rebalancing into. At 10.3934% for five years and 11.2195% for ten, against 4.2% inflation, the long end of the government's own curve is paying a real yield of roughly six to seven points. The Ghost is not reaching for yield outside the sleeve it already owns; it is buying more of the same government paper further along the same curve.

1.5m to Treasury bills, at the 364-day rung. This is a correction to last month's reasoning, published as one. August moved the whole bill allocation to bonds because the 91-day was clearing at 3.35% against 4.0% inflation. Auction 1205 shows the 91-day still below inflation at 3.3961%, so that half stands. But the same auction cleared the 364-day at 6.6492%, a positive real yield, and August's note never separated the two. The instrument was never the problem. The rung was. The bill sleeve is fed again, one year out rather than three months out, which also keeps a rolling liquidity ladder alive underneath the bonds.

1.0m to cash. Cash is below the middle of its band, and there is a specific reason to keep dry powder this month rather than commit it. The index series carries an unexplained scale discontinuity dated 26 August, and no verified 25 August equity session exists in the snapshot the Ghost reads. Until the exchange explains it, week-over-week index returns cannot be computed. Holding a little more cash is the honest response to a tape you cannot yet measure.

0.5m to the gold sleeve. A small proportional top-up, nothing more. The commodity sleeve sits at 3.7% against the middle of its 0 to 8% band, so as the book grows the hedge has to grow with it or it quietly shrinks. The Bank of Tanzania's July Monthly Economic Review recorded gold exports up 36.4% year on year to US$5.5229bn, which is the reserve-accumulation backdrop this sleeve was opened against in June. The June lesson stands: a hedge is bought early, or not at all.

And nothing to equities, for the second month running. Repeating a decision is not the same as forgetting to make one, so it is published again. Three reasons. The sleeve is the largest in the book and the furthest above its band middle. It is concentrated: one holding is roughly a fifth of the entire book, and on 28 August that same counter carried 83.8% of the exchange's traded value across 20 counters, of which 7 advanced and 9 declined. And it is currently the least measurable sleeve the Ghost owns, because of the 26 August discontinuity. The Ghost sold nothing. It let the other sleeves grow instead, which walks equities from 54.7% toward 54.1% without a single trade.

Model-book context: how The Ghost reads this terrain → Fixed income · Commodities · Equities · Government securities board

After the allocation every sleeve remains inside its mandate band: equities 54.1%, fixed income 32.4%, cash 9.8% and commodities 3.7%. No band-breach rationale is required. Nothing was sold. This is new savings finding its place, not a rotation.

The book is larger this month because the Ghost saved, not because it earned. Contributed capital rises to TSh 521.0m while market P&L is unchanged at TSh 67.4m, and the money-weighted return holds at +13.28%, exactly as it should, because a contribution landing at period end carries no weight in a modified Dietz calculation. Had the Ghost divided the new book value by the original seed instead, it would have printed +17.67%. That four-point gap is entirely savings, and reporting it as performance would be a lie the arithmetic makes easy. The two figures sit side by side on the tracker so the difference is always legible.

Two further honesties. The 5.5m added to the fixed-income sleeve has no history in this book yet, so it will appear in future bridges rather than retroactively in past ones. And the holdings behind these figures are marked at the 31 August tracker mark, which itself prices equities at the last verified DSE close of 28 August 2026, because no verified session existed for 31 August when this allocation was made. No index return spanning the 26 August discontinuity is quoted here, and none will be until the exchange explains it. null > fabricated, always.