Work that compounds.
Shabba builds algorithmic market-making infrastructure for East African capital markets. The systems we build run for years. The decisions matter for decades. If that is the kind of work you want to do, the rest of this page is for you.
Your first year, concretely.
Four things we screen for, regardless of background.
Substance over signal
A pull-request you can defend in detail beats a CV with brand-name logos. Show the work, code, papers, models, post-mortems, anything specific.
Long horizon over fast iteration
The systems we build run for years. We optimise for decisions whose payoff arrives years after the work is done, not the next quarter.
Discipline over cleverness
A boring strategy executed cleanly outperforms a brilliant strategy executed sloppily. Risk frameworks before alpha-hunting.
Direct over diplomatic
The right answer wins regardless of who said it first or how senior they are. Strong opinions, well held, openly debated.
The bar is not you already know all of this. The bar is you can learn this faster and more rigorously than most. The strongest applications we get are from people whose current portfolio doesn't cover everything we do, but who can show, in detail, how they have learned hard things from scratch before. If that sounds like you, write to us.
Three tracks, one operating standard.
We hire against the substance of the work, not against rigid job titles. If the description below sounds like what you want to do, write to us.
Quant.
Own the pricing engine and the risk system. Black-76 for futures-style options. SVI surfaces calibrated continuously against live prints. DV01, Greeks, basis tracking across cash, futures, and options legs. The work is half mathematical (deriving the right model for the right instrument), half operational (running it against real markets where the data is sparse and the regime shifts).
Engineering.
Build the order management and exchange gateway layer. FIX and native protocols against DSE, NSE, and the venues we extend into. The data plane the quants depend on: tick captures, calibration pipelines, model deployment to live. The kind of work where a bad p99 is treated as a defect, not a feature request.
Operations.
Own the regulator and exchange relationships across the markets we operate in. CMSA, CMA Kenya, SEC Nigeria, NSE, DSE. The work of getting and keeping the licences, responding to consultation processes substantively, and translating between technical reality and policy framework. This track is small and senior.
Problems we actually think about.
One per track. If any of these are the kind of problem you want to spend your days on, you are probably the person we want to talk to.
The hedge-slicing trade-off.
You have written 100 lots of an at-the-money one-month option on a name whose underlying Single Stock Future trades 200 lots per day on average. The position has a delta of 50, so you must buy 50 SSF contracts to be neutral. You can slice that 50-lot hedge into N equal child orders over the remaining trading hours of the day.
At N = 1, you take 25% of daily print volume in a single market order. Expensive in market impact, but you are instantly hedged. At N = ∞, you carry delta-drift risk until the last child fills, but pay zero impact at the margin.
Derive the optimal N as a function of per-unit impact cost c, delta-drift volatility σ, and time horizon T. State your assumptions explicitly.
10 weeks. Real work. Same standard.
A summer programme on the same three tracks the firm builds around. The bar for what you ship is the bar full-time staff are held to. We hire from the intern pool ahead of new-grad hiring.
Ready to write?
Send a short note about what you want to work on, a CV, and one substantive piece of work. Address it to careers@shabba.tz or use the contact form.