Two LuSE stocks whose screen prices tell a misleading story. ZAFFICO's fall is mostly its rights issue showing up in the price. Copperbelt Energy's fall is partly the kwacha getting stronger, because CEC earns in US dollars.
ZAFFICO grows pine and eucalyptus plantations and is moving into wood manufacturing. In July it offered every shareholder one new share for each share held, at a deep discount, to fund that plan.
For a one-for-one offer, the share should settle near the average of the old price and the offer price.
Through August the shares still changed hands around , roughly above that level, even though buyers no longer received the rights. Trading is thin, so the price only caught up in September. The latest price, , is that level.
The take-up result was not published where we could find it, so today's share count is a range. Uses full-year 2025 earnings per share.
| Scenario | Price | Shares | Value | P/E |
|---|
CEC supplies power, mostly to copper mines in Zambia and the DRC. It reports in US dollars, but its shares trade in kwacha, so a stronger kwacha shrinks the kwacha value of the same dollar profits.
From the start of 2026 to the latest price. Exchange rate from EODHD closes as reported by The Rio Times.
Much of the kwacha fall is the currency, not the business. In dollar terms the drop is much smaller, while dollar revenue rose and profit held flat in the first half.
Live price against full-year 2025 earnings, and first-half 2026 results.
Every figure in this section is recalculated from the price data each time the page loads, using the same steps for both companies. Each price move is split into known effects, the market and a company-specific remainder, and the remainder is then tested against the stock's normal day-to-day noise. Each finding is labelled with how much weight it can bear.
Estimates that need more data fill in automatically as the Worker collects a price every trading day.
| Measure | ZAFFICO | Copperbelt Energy |
|---|
Likely range assumes recent volatility continues and no drift. It describes uncertainty, not a forecast of direction.
| Likely range | ZAFFICO | Copperbelt Energy |
|---|
Log price changes add up, so a move can be split exactly. For ZAFFICO, from the last close before the rights issue: the dilution from issuing one share at K2.53 per old share (log of expected price after ÷ price before), the market, and a remainder. For Copperbelt Energy, from 1 January: the change in kwacha per dollar (CEC reports in dollars, so a stronger kwacha lowers its kwacha value one-for-one), the market, and a remainder.
With at least 20 matched observations, the market's share uses a Dimson (1979) beta: the stock's returns are regressed on the LASI's returns over the same period, the period before and the period after, and the three betas are summed. This corrects the bias thin trading puts on an ordinary beta. Observations are weighted by the number of trading days they cover. With less data, beta is set to 1 (the market-adjusted model, Brown and Warner 1985).
The company-specific remainder is divided by the stock's volatility scaled to the length of the window (z = remainder ÷ (daily volatility × √trading days)), and compared with a t distribution. Daily volatility is estimated from irregularly spaced prices, assuming variance grows with trading days. A big remainder that isn't significant means the stock is too volatile, or the data too thin, to call it more than noise.
A discrete Ornstein–Uhlenbeck regression on the log gap between the price and the theoretical ex-rights price, using prices since the ex-date: change in gap = −κ × trading days × previous gap. A significant κ > 0 means gaps close over time, with half-life ln 2 ÷ κ. The expected price assumes full take-up of the new shares.
Once 20 matched days are collected, CEC's daily returns are regressed on the change in kwacha per dollar and on the LASI. An elasticity of 1 means the market prices CEC as a dollar asset; 0 means no currency link. Both are tested.
Before daily collection builds up, history comes from about 20 hand-collected closing prices per stock, so most tests have little power. LuSE holidays are treated as trading days. Thin trading makes prices stale, which understates volatility for stocks like CEC. Collected days where price and volume are both unchanged from the day before are treated as no fresh data. Nothing here is investment advice.