Diversification, and the part of risk it cannot touch

Diversification is the one idea in investing that gives something for nothing, and it is routinely oversold. It works on a specific kind of risk and is powerless against another kind, and most disappointment comes from expecting it to do the second job.

Two kinds of risk

Specific risk belongs to one asset: this building has a structural problem, this tenant left, this street was rezoned, this developer failed. Spread across many assets and these average out, because they are unrelated to each other. Systematic risk belongs to the whole market: interest rates, the economy, national policy, security events. Every asset carries it, so owning more of them does not reduce it. Diversification is a tool for the first kind only.

What this means in one country

Israeli residential property in different cities is genuinely less correlated than the same street twice, but not nearly as uncorrelated as it looks. The same central-bank rate, the same mortgage rules, the same land-supply policy, the same tax regime and the same national events reach all of it. Holding assets in several cities is worth doing and is not the same as being diversified — a national shock moves them together.

The dimensions that actually differ

Geography helps a little. More useful is spreading across things that fail for different reasons: an income-producing standing asset alongside a development, which are exposed to tenants and to construction respectively; different hold periods, so that not everything needs to be sold into the same market; different tenant types; different price points, since the cheap end and the expensive end of a market do not always move together. And beyond property entirely — a portfolio of only property is concentrated no matter how many buildings are in it.

The trade-off nobody mentions

Splitting a fixed budget over more positions means smaller tickets, and smaller tickets mean any fixed or minimum fee weighs more heavily as a proportion. There is also an attention cost: ten positions you have not read about are worse than three you have. The benefit of adding positions also falls away quickly — going from one asset to four removes most of the single-asset risk, and going from twelve to sixteen changes very little.

Time is a dimension too

Committing everything at one moment is a concentrated bet on that moment's prices. Spreading commitments across time averages the entry point. In an illiquid asset class where you cannot easily correct a badly timed entry, this matters more than it does in liquid markets.

What it does not do

It does not turn a set of weak assets into a good portfolio; averaging poor decisions produces a reliably poor result. It does not create liquidity — twelve illiquid positions are still illiquid. It does not protect against a market-wide fall. And it cannot be relied upon in the moment it is most wanted, because correlations rise in a crisis: the assets that behaved independently for years tend to fall together when the cause is common to all of them.

The honest summary

Diversify because avoidable concentration is a needless risk, not because it makes property safe. What remains after diversifying is the risk of the asset class itself, and that risk is the reason there is a return in the first place.