Diversification in Pakistan: what it can and cannot do in a one-market portfolio

KARACHI: "Do not put all your eggs in one basket" is the oldest sentence in investing, and like most old sentences it is both true and incomplete. Diversification genuinely protects you from some risks — the factory fire, the fraud, the single bad management team — and offers almost no protection against others, such as a currency crisis that drags down every share on the exchange at once. Pakistani investors face a sharper version of this truth than most: one stock exchange, one currency, one central bank, and an index dominated by a handful of sectors. Understanding exactly which risks diversification removes, which it merely softens, and which it cannot touch is the difference between a portfolio that is actually diversified and one that is merely long. This guide works through the logic with the local market's real shape in mind.

The one risk diversification truly eliminates

Statisticians divide risk into two kinds. Unsystematic risk belongs to a single company: its plant floods, its chief executive resigns in scandal, its main product is banned, its biggest customer defects. Systematic risk belongs to the whole system: interest rates, the exchange rate, politics, the world oil price. Diversification — holding many companies instead of one — attacks the first kind almost completely and the second kind barely at all. That asymmetry is the whole story of this article, so let it land with an example.

Imagine, illustratively, two investors with Rs 1,000,000 each. The first holds one mid-sized company. The second holds twenty companies of Rs 50,000 each, across banking, fertilizer, cement, pharmaceuticals, food, energy and technology. Now suppose one company — a different one in each portfolio — suffers a disaster and loses 80 percent of its value. The first investor loses 80 percent of everything: Rs 800,000 gone, because their one egg was the one that broke. The second investor loses 80 percent of one-twentieth: Rs 40,000, or 4 percent of the portfolio — painful, forgettable within a year or two of ordinary returns. Nothing about the second investor's forecasting was better. They did not predict the disaster; they made its size irrelevant. This is why diversification is sometimes called the only free lunch in finance: it reduces the damage from individual failures without requiring you to know which company will fail.

Research across markets consistently finds that most of this benefit arrives early. Moving from one holding to five removes a large share of single-company risk; by fifteen to twenty reasonably unrelated holdings, the remaining unsystematic risk is small, and the fiftieth holding adds paperwork rather than protection — a point our step-by-step guide to building a PSX portfolio echoes when it suggests keeping holdings to a number you can actually follow. Beyond that point, extra safety must come from a different direction entirely: different asset classes, as discussed below.

Correlation: the word that decides everything

Diversification works only to the extent that your holdings do not move together. The statistical name for togetherness is correlation: two shares that always rise and fall in step are, economically, one holding with two names. This is where nominal diversification deceives. Five Pakistani banks are not five independent bets; they share the same regulator, the same interest-rate cycle, the same government-securities market and largely the same deposit base. When the State Bank cuts rates or a banking-specific shock hits, they will mostly move as one. The same is true of five cement companies (one construction cycle, one coal price), or five textile exporters (one cotton crop, one exchange rate, one set of foreign buyers).

Genuine diversification, then, is measured in independent drivers, not in ticker symbols. Ask of each holding: what has to go right for this to do well, and what has to go wrong for it to do badly? A fertilizer producer's fortunes turn on gas supply and pricing, farm incomes and urea demand. A pharmaceutical company's turn on drug-pricing decisions and input costs. An oil explorer's turn on international crude prices and the dollar — its earnings often rise in rupee terms when the currency weakens, which can partially offset holdings that suffer from the same depreciation. A software exporter earns dollars and spends rupees, another partially offsetting profile. A portfolio assembled from businesses with different drivers will still fall in a general crash — everything does — but its recoveries, dividends and day-to-day behaviour will be far less hostage to any single policy decision or commodity price.

A simple home test: look at your holdings and group them by "what hurts this?" If one event — say, a sharp rise in interest rates — lands in almost every group's answer, you are concentrated in that factor whatever the sector labels say. Interest-rate sensitivity is the most common hidden concentration in Pakistan, because high policy rates simultaneously squeeze leveraged manufacturers, reprice bank earnings, raise the government's borrowing costs and pull money towards risk-free instruments. Noticing the factor is most of the cure.

The Pakistan problem: one market, one currency

Now the honest limits. A PSX-only portfolio, however cleverly spread across sectors, remains exposed to Pakistan-level risks that no amount of local spreading removes: a balance-of-payments crisis, a steep rupee depreciation, an IMF programme's conditions, political instability, or a domestic rate shock. In severe episodes — and Pakistan has had several, documented in our history of PSX crashes and their lessons — correlations across local shares rise towards one: in a true panic, investors sell what they can, not what they should, and good companies fall with bad ones. Diversification within the market smooths ordinary years; it does not insure against the country's bad years.

Investors respond to this at three levels, and it is worth separating them. The first is asset-class diversification at home: holding some wealth outside equities altogether — in government securities or money-market funds, in gold, in property, in a business, in cash. These assets respond to Pakistani shocks differently from shares (some, like gold priced in dollars, have historically moved opposite to the rupee), so the household's total position is steadier than any single part. The second is income diversification: a salary, a business or remittances from family abroad are also "holdings" in the broad sense, and a family whose income arrives in dollars from overseas is naturally hedged in a way a purely domestic earner is not — one reason the Roshan Digital Account channel matters to overseas Pakistanis investing back home. The third level, geographic diversification into foreign markets, exists but is heavily constrained for residents by State Bank rules on taking capital out; it is mostly relevant to Pakistanis already based abroad. For most readers, the practical frontier is level one: the split between shares and non-share assets, which is exactly the allocation decision our portfolio-building guide treats as the master choice.

Concentration inside the index itself

Investors who buy "the market" through an index product sometimes assume the concentration problem has been solved for them. On the PSX it has only been softened. The KSE-100 is weighted by free-float market capitalisation, so its largest members — big banks, energy and fertilizer names — account for a disproportionate share of the index's movement, and a few sectors dominate its character. Holding the index plus six bank shares is not broader than holding the index; it is the index with extra banking. None of this makes index exposure a poor choice — it is diversified against single-company disasters by construction — but read its sector weights once, so you know which Pakistan you actually own. Our explainers on PSX indices beyond the KSE-100 and ETFs available locally fill in the product details.

The mirror-image trap is diworsification: adding holdings not because they diversify anything but because they were tipped, cheap-looking or already owned by a friend, until the portfolio is a museum of other people's decisions — forty positions, several in the same sector, none large enough to matter, all demanding attention at results time. Diversification has an optimum, not a maximum. Every holding should be able to answer the question "what risk do you offset, or what driver do you add?" in one sentence.

Diversification across time

There is one more axis, easy to overlook because it involves no new securities at all: time. Buying your whole position on one day concentrates your entry price in one market mood; spreading purchases across months and years diversifies the prices you pay, so that neither a euphoric peak nor a panic trough defines your cost base. This is the principle formalised by rupee-cost averaging — investing a fixed amount on salary day regardless of headlines. Time diversification does not raise expected returns, and it will underperform a lump sum in a market that rises steadily from the day you start. What it buys is insurance against the single worst start date, and — just as valuable for most people — a routine that keeps investing happening automatically through the scary months when discretionary buyers freeze.

Rebalancing interacts with all of this. As winners grow, a diversified portfolio quietly concentrates into whatever performed best, which is usually whatever is most exposed to the most recent favourable factor. The periodic rebalancing described in our portfolio guide is, in effect, diversification maintenance: it keeps the spread of drivers you originally chose from eroding into a bet you never consciously made.

A practical diversification audit

Twice a year, lay your CDC statement out and ask five questions. One: what is my largest single holding as a percentage of the portfolio, and could I shrug off its halving? Two: what is my largest sector, and do I actually have a view on it, or did it just accumulate? Three: list each holding's main driver — rates, oil, the dollar, construction, consumer spending — and count how many share the same driver. Four: what share of my total wealth (not just the brokerage account) sits outside shares entirely? Five: when did I buy, and is my cost base hostage to one period? Where the answers embarrass you, fix slowly with new savings first — directing fresh money to underweight areas restores spread without triggering spreads, commissions and taxes on sales. Aktzien is being built with exactly this kind of small, regular investing in mind — it plans to offer fractional shares, so a modest monthly amount could be split across several companies instead of concentrated in whichever single share the full amount happens to afford — but the audit above works today, with any broker, on paper.

Hold the balanced conclusion. Diversification is powerful against the risks that strike one company at a time, helpful against sector shocks, weak against country shocks, and useless — by itself — against your own behaviour in a panic. That is not an argument against it; it is an argument for pairing it with the other defences this series covers: sensible position sizes, an emergency fund outside the market, and position-level risk rules. This article is educational content, not financial advice; how much concentration you accept is a personal decision that should fit your circumstances and your honest tolerance for loss.

A worked picture of spread versus concentration

Make the idea concrete with an illustrative comparison. Two portfolios of Rs 1,000,000 each face the same year: interest rates rise sharply, construction slows, but the rupee's slide lifts exporters and dollar earners. The first portfolio holds four cement and steel companies — one driver, effectively. The rate rise squeezes their financing costs while the construction slowdown cuts volumes, and the portfolio ends the year down heavily, with every holding telling the same sad story in the same quarter. The second portfolio holds a bank, a fertilizer producer, a pharmaceutical company, an oil explorer, a food producer, a software exporter and two others besides, with the rest in a money-market fund. The same year hits its construction-linked holdings, but the explorer's dollar-linked earnings translate into more rupees, the exporter's order book firms, the bank's interest income reprices, and the money-market sleeve yields more every month as rates climb. The second portfolio may still finish the year down — diversification does not repeal bad years — but it finishes down by a fraction, with dividends still arriving from several directions and no single policy decision able to wound every holding at once. That gap in lived experience, repeated over decades, is the entire prize: not higher returns in the good years, but a portfolio whose bad years are survivable without panic, forced selling or abandoned plans.

Key takeaway: diversification reliably neutralises single-company disasters and softens sector shocks, but in a one-market, one-currency portfolio it cannot repeal Pakistan-level risk — that part is managed by asset allocation, time and position size, not by owning more tickers.