Should you split one sum across products?
Nobody argues with “do not put all your eggs in one basket”, and then it gets vague fast. Into how many parts? Split along what? When is it enough? More to the point — what does splitting actually protect against, and what does it leave untouched?
Answering the last one first, because it decides whether the others are worth any effort.
What spreading reduces, and what it does not
The Wikipedia entry on diversification in finance separates risk into two kinds: one that is diversifiable and can be reduced by spreading holdings, and one that is non-diversifiable — which, in the entry’s own words, exists however many holdings you buy.
Carry that over to yield products. Spreading reduces the risk that one product, one issuer or one route fails on its own. It does nothing about the market falling as a whole — five different products all holding the same volatile asset go down together.
This gets misapplied constantly. Splitting a sum across five products feels like risk management, but if all five sit on the same underlying, the split addressed the first kind of risk and not the second at all. What you diversify is not the number of parts; it is the source of risk.
Three dimensions worth spreading along
- Product. Flexible, fixed-term, staking-type structures fail in different ways, so separating them stops one mechanism’s problem taking everything.
- Term. Staggering maturities. This mostly does not address losses; it addresses “needing the money and none of it being reachable”. It is the same subject as what a lock-up costs, seen from the other side.
- Platform. Different custodians. This is the dimension that maps to issuer or platform credit risk, and the only one of the three that reduces counterparty exposure.
They are not equally worthwhile. Term is almost always worth doing — lowest cost, most concrete problem solved. Product depends on whether you have enough to spread without thinning it out. Platform costs the most: another account to secure, another set of records, another interface to learn. That makes it a larger-balance decision.
The cost of overdoing it
More is not better. Spreading carries three real costs:
- Attention. Every extra place is another set of rules, maturity dates and changes to track. Past your capacity, you become more likely to miss a notice, not less.
- Operational error. Moving funds around is itself exposure — wrong field, wrong destination, forgetting to withdraw. Those are risks the spreading created.
- Thinning out. Where a product has rate tiers, splitting too finely can land every slice in a lower tier and reduce the total.
A rough but effective test: if you cannot say where each part is, when it matures and who to contact if it goes wrong, you have split too far. The number you can hold in your head is the number you actually control.
A workable order
You do not have to do all three at once:
- First, sort into layers. Whether the money can sit still, and for how long, matters far more than how many parts it becomes — see sort your money into layers first.
- Second, stagger maturities within the layer that can be locked. Cheapest step, most direct payoff.
- Third, if the amount has reached the point where keeping it all in one place would bother you, consider the platform dimension.
- Fourth, product last. It is mostly an adjustment to the shape of returns rather than a risk necessity.
The logic of the order: deal first with the problem most likely to actually happen (needing money you cannot reach), then with the one that hurts most but happens rarely (one place failing). Reversed, people usually get caught by the first kind.
When not to spread
At small amounts the benefit is usually below the cost. Splitting a modest sum three ways removes little risk and adds a real management burden. The effort is better spent deciding which layer that single amount belongs to.
Another case: the parts you split into all rest on the same underlying. That buys reassurance rather than protection — as above, it does nothing for the non-diversifiable kind.
And the last one: reaching for a product or platform you do not understand in order to be more spread out. That is not diversifying risk; it is adding a kind you cannot see. Spreading assumes you can read every part you hold.
This is a risk-management framework. It recommends no specific product or platform combination and is not investment advice. Spreading cannot remove losses caused by broad market movements; product rules, rate tiers and exit conditions are whatever the official page says when you open it. Yield products can lose principal.
Common questions
How many parts is right?
No standard number. A usable ceiling: can you say, without checking records, where each part is, when it matures and who to contact if something breaks? If not, you have split too far. The floor is set by size — at small amounts the added management burden usually outweighs the benefit.
Does spreading protect against a market fall?
No. The Wikipedia entry on diversification separates diversifiable from non-diversifiable risk and notes that the latter exists however many holdings you buy. If your several products all sit on the same volatile underlying, they fall together. Spreading addresses one place failing on its own, not the market as a whole.
Which dimension should I do first?
Sort into layers, then stagger maturities. Layering addresses the problem most likely to actually occur — needing money you cannot reach — and staggering costs the least. Platform is the most expensive and suits balances large enough that one location would bother you. Product goes last; it shapes returns more than it reduces risk.