Diversification, fees and time horizon
Three things that decide how an investment actually turns out: not putting everything on one event, keeping the certain costs low, and matching what you hold to when you need it.
Diversification
Spreading money across things that would not all fail for the same reason. The point isn't owning many things. It's owning things whose bad years are caused by different events. Five companies in one industry, in one city, selling to the same customers, is one bet written out five times.
The test is a question you ask out loud: name the single event that would hurt all of these at once. If it comes to you easily, you are not diversified against it.
One field or five
A farmer with one crop in one field has a wonderful year when the weather suits that crop and nothing at all when it doesn't. Five crops in five valleys means no spectacular year and no empty one. Where the analogy breaks: some events reach every field at once, like a drought across the whole region or a currency that collapses. Diversification softens the specific disasters, not the general ones, and nothing removes them all.
Check yourself
Omid says he is well diversified: his money is split between five different companies, all of them builders working in the same city. What is the problem?
- A single event, a slump in that city's building work, would hit all five at once
- Five is too few. The number really needs to be closer to fifty before it counts
- Nothing at all: holding five separate companies is genuine diversification
- He should hold one company instead and follow it closely enough to sell in time
Show the answer
A single event, a slump in that city's building work, would hit all five at once
Right. He owns five names and one bet. Diversification is about the reasons things fail, not the number of things owned.
Why a big fall is worse than it looks
Everything in one thing, and it falls 60%
100 becomes 40
to get back to 100: 100 / 40 = 2.5
so it has to rise +150%
Spread across five, and one of them falls 60%
five lots of 20; one of them becomes 8
8 + 20 + 20 + 20 + 20 = 88
to get back to 100: 100 / 88 = 1.136
so it has to rise +13.6%Same disaster, same 60% fall.
One bet: a 150% climb back. Spread: a 13.6% climb back.Losses are not symmetrical. A 50% fall needs a 100% rise to undo, and a 90% fall needs a 900% rise. Avoiding the catastrophe matters more than catching the best year.
Check yourself
Spreading money across many different things removes the risk of losing money.
Show the answer
False
It removes the risk tied to any one of them, not the risk of the whole lot falling together. Years where nearly everything drops at once have happened many times, in many countries. Diversification lowers your chance of a catastrophe and, in the very same move, lowers your chance of a spectacular year. That trade is the whole deal.
What it does, and what it doesn't
- It removes the risk specific to one thing: one company failing, one tenant leaving, one building burning down.
- It does not remove the risk of everything in the same economy falling in the same year.
- It lowers your chance of a disaster and your chance of a spectacular year together. You cannot buy one without the other.
- Spreading across different kinds of asset, and where you can, different countries and currencies, does far more than owning many of the same kind.
- Past a point, adding more of the same kind changes almost nothing. The first few steps do most of the work.
THE CERTAIN COST
Fees are the number you can actually control
Returns are uncertain. Costs are not. A yearly fee is taken whether the year was good or bad, and it comes out of the total, so it also removes everything that money would have earned later. A small percentage charged every year compounds against you in exactly the way growth compounds for you.
Fees arrive under many names: management charge, commission, spread, platform fee, the gap between the buying and the selling price. The question is always the same one. What percentage of my money leaves every year?
10,000 held for 20 years, growing 8% a year before costs
(8% is for illustration only, not a promise)
With no yearly fee
10,000 x 1.08^20 = 46,610
With a 2% yearly fee, so 6% left over
10,000 x 1.06^20 = 32,071No fee 46,610
2% a year 32,071
The fee took 14,539 - more than the 10,000 you started withNothing about the underlying thing changed. Two percent a year, taken quietly, removed close to a third of the final amount.
Check yourself
Mina is choosing between two options that hold much the same things. One charges 0.5% a year, the other 2.5%. The expensive one has done better over the last two years. What should she take from that?
- The extra 2% is clearly worth paying, since it is already earning itself back
- She should pick whichever of them had the single best year
- Fees are too small to matter next to the size of the returns involved
- Two years says little about future returns; the 2% gap is certain every year
Show the answer
Two years says little about future returns; the 2% gap is certain every year
Yes. The cost is known in advance and repeats every year forever. Two years of relative performance is mostly noise. The certain number deserves more weight than the uncertain one.
Time horizon
When you will need this particular money. It is the most useful thing to know before deciding how it should be held, because it sets how much of a fall you could ride out without being forced to sell. Money with a short horizon cannot take swings, however good the long-run case for something is.
The same person holds several horizons at once: rent due next month, a car needed in three years, money for a life stage decades away. Those are three different questions, not one.
Check yourself
How far away is each of these?
- Next month's rent and this season's bills
- The emergency fund, which might be needed tomorrow
- A car you plan to replace in three years
- A wedding you are saving for, two years out
- Money for a life stage twenty years away
- A child's education starting in twelve years
Show the answer
Within a year: Next month's rent and this season's bills, The emergency fund, which might be needed tomorrow
One to five years: A car you plan to replace in three years, A wedding you are saving for, two years out
Five years or more: Money for a life stage twenty years away, A child's education starting in twelve years
Check yourself
Match each situation to what it really means
Show the answer
- Five companies in the same industry → One bet written out five times
- A 60% fall → Needs a 150% rise just to get back
- 2% a year in fees over 20 years → Roughly a third of the final amount
- Money needed next month → A horizon too short for any swing
- No emergency fund → What turns a fall into a permanent loss
Lesson recap
- Diversification is about the reasons things fail, not the number of things you own. Name the event that would hit them all.
- Big falls are not symmetrical: down 60% needs up 150%. Spreading out is mostly about avoiding that hole.
- Fees are certain while returns are not. 2% a year over twenty years took about a third of the final amount.
- Your time horizon decides what the money can be held in. Short-horizon money cannot take swings.
- The forced seller does the real damage, and an emergency fund is what stops you becoming one.