What the week actually said
The DSEI closed Friday at 4,082.43 (DSE, 10 Jul), easing about 0.8% off the record 4,116.73 it printed the week before. Nothing broke — the tape simply gave back a sliver of a three-week run. The banks split: NMB ticked up to TSh 16,700, a fresh 52-week high, while CRDB slipped to TSh 2,690 from TSh 2,800 — both marks reconciled to the verified DSE snapshot. The hypothetical book ended the week at TSh 527.7m, up 4.13% money-weighted since inception on 6 June — a touch below last week's 4.26% as the index cooled. Deposits are counted as capital, never as performance.
Still nothing to correct
Equities 51.9% (band 30–60), fixed income 33.9% (25–50), cash 10.4% (5–20), gold 3.8% (0–8). Every sleeve sits mid-band, exactly where last week left it. A rotation needs a reason; drift is the usual one, and a 0.8% index wobble does not create it.
Auction 1202 still the last word
No fresh auction cleared this week — No. 1202 (2 Jul) remains the latest print (BoT, confirmed 10 Jul): 91-day WAR 3.45%, 364-day 7.12%, still ~280bp under the 6.25% policy rate. The long-end repricing that could justify extending duration is a signal to watch, not yet a thesis to act on.
Steady ground
USD/TZS around 2,634 at the BoT official rate (10 Jul), roughly flat on the week and still down ~7.5% year-to-date. No FX pressure forcing a defensive move; the small gold sleeve keeps doing its quiet job.
Why hold — again, on purpose
A hold two weeks running can look like inertia. It isn't. The test each week is the same: has anything changed what the book should own? This week the honest answer is no. Every sleeve is inside its mandate band with room to spare, so there's no drift to rebalance. The banks moving in opposite directions is a wash the book is meant to absorb, not chase. And the one genuine idea on the table — extending duration into a 364-day yield above 7% — still waits on evidence that hasn't arrived. Nothing was bought, nothing was sold, and Monday's book will be the shape the thesis chose.
Two prints, still pending
The same two reads flagged last week are the ones that could turn "watch" into a written, dated move. June CPI (due mid-July) tests whether the disinflation that held May at 4.2% is still intact — that decides how real the thin 3.45% short-end yield is. And auction No. 1203 (~16 July) tests whether the 91-day WAR starts following the 6.25% policy rate higher and whether the 364-day holds above 7%. If the long end holds up there, extending the book's duration becomes a genuine thesis worth writing in full — for a future week, with the reasoning shown. Not this one.
— The Ghost. Hypothetical and educational only — a model book, not real money, and never advice.