For example, our $ABC stock has a 10% chance of a 50% drawdown. And the other 90% of the time, the stock grows 10% per year.
Plugging these assumptions into the Kelly Criterion, we find that the *optimal* portfolio is 20% cash and 80% $ABC stock.
Calculations:
10/
So, when $ABC stock is 100% of our portfolio, our long-run CAGR is ~1.66%.
How much does this CAGR increase if we follow the Kelly Criterion -- ie, 80% $ABC stock, 20% cash, and annual re-balancing to stay that way?
Well, this improves our CAGR from ~1.66% to ~1.83%.
11/
All that fancy math and effort -- for a measly 0.17%.
Where's the justice in this universe?
And furthermore, Kelly tells us that this portfolio is *optimal*.
So, NO other "cash + $ABC + re-balancing" portfolio can beat this ~1.83%/yr either.
12/
So, IF we want to escape the effect of a big drawdown, our portfolio cannot have just cash and $ABC stock in it.
We need a new asset.
And this asset can come from a rather unlikely place.
13/
Suppose we have a second stock, $XYZ.
Unlike $ABC (which goes *up* most years), $XYZ goes *down* most years. And not by a small amount either -- but by 25%.
*Occasionally* though (about once in 10 years), $XYZ does super well. It 10-Xs -- ie, goes up +900%.
14/
$XYZ, it turns out, is a much *worse* stock to buy and hold than $ABC.
With $ABC, we at least get a *positive* return: ~1.66% per year.
But if we buy and hold $XYZ, we end up with a *negative* ~2.82% long-run CAGR.
One 10-X does NOT make up for nine 25% drops.
15/
So, over the long run, $XYZ LOSES money.
Why even bother with $XYZ then?
Because it has one redeeming feature: it's *negatively correlated* with $ABC.
Remember the 1 year out of 10 when $ABC experiences a 50% drawdown? That happens to be the year $XYZ 10-Xs.
16/
We can take advantage of this negative correlation.
How? We build a portfolio containing *both* $ABC and $XYZ.
9 out of 10 years, the $ABC part of the portfolio grows 10%. And the $XYZ part shrinks 25%.
When that happens, we sell some $ABC to buy *more* of the LOSER $XYZ.
17/
But 1 year out of 10, $ABC will experience a 50% drawdown.
That same year, $XYZ will 10-X.
And that's when we sell part of our $XYZ holdings to buy $ABC on the cheap.
Negative correlation + Re-balancing is a powerful combination.
18/
The Kelly Criterion again helps us work out the optimal split between $ABC and $XYZ.
It turns out to be: ~73.31% $ABC and ~26.69% $XYZ.
And this optimal portfolio achieves a ~12.41% per year long-run CAGR -- with NO drawdowns in ANY year!
19/
So, we have:
Just $ABC: ~1.66% CAGR, 50% drawdown,
Just $XYZ: *negative* ~2.82% CAGR, 25% drawdown,
$ABC + Cash: ~1.83% CAGR, 40% drawdown, and
$ABC + $XYZ: ~12.41% CAGR, NO drawdown.
That's the power of negative correlations and re-balancing.
20/
Of course, in the real world, stocks produce a range of returns -- not just +x% or -y%.
And the probabilities of these returns are not known so precisely.
And perfect negative correlations are hard to find.
So, our example wasn't super-realistic.
21/
Still, there are some broad lessons we can take away from this exercise.
First, we should look at investments from a "portfolio" perspective.
*Standalone*, a bet may have negative CAGR. But it may still add much value to a *portfolio* of other bets (eg, our $XYZ stock).
22/
Second, negatively correlations provide powerful diversification.
We can't usually predict the future.
But with negative correlations, we *can* construct our portfolio so that: no matter how the future plays out, there are almost always *some* bets that are paying off.
23/
The bets that are paying off provide liquidity and minimize portfolio level drawdowns.
This can give us opportunities to rotate out of them and re-balance into other assets on the cheap.
If we do this successfully, we can meaningfully improve our portfolio's performance.
24/
A financial advisor named Harry Browne popularized these ideas in the 1970s.
Browne suggested that investors construct their portfolios in a "future agnostic" way.
Whether the future brings growth or recession, inflation or deflation, the portfolio should still perform:
25/
Several well-known investors were inspired by Browne.
They developed Browne's ideas further and added their own unique spins.
For example, @RayDalio with his All Weather Portfolio, @MebFaber with his Trinity Portfolio, Mark Spitznagel with his Tail Hedging strategies, etc.
26/
My friend @TaylorPearsonMe and his colleague @JasonMutiny have also modeled their investment philosophy along similar lines.
Their approach is to generate negative correlations via "volatility" strategies (eg, VIX futures).
They call this the Cockroach Portfolio.
27/
These "negative correlation" techniques are super fun, and the core concepts behind them can be useful to investors of all stripes.
To learn more, please join @TaylorPearsonMe, @JasonMutiny, and me tomorrow (Sun, Mar 27) at 1pm ET on Money Concepts:
In this thread, I'll walk you through the art and science of valuing a company.
The Goal: We want to figure out how much a company is worth -- so we know what's a good price to pay for its shares.
2/
Warren Buffett often says that ALL intelligent investing is value investing.
What exactly is value investing?
Many people think value investing means buying companies at low P/E (Price To Earnings) ratios, or low P/B (Price To Book) ratios, or high dividend yields.
3/
But that's NOT how Buffett looks at it.
To Buffett, value investing simply means buying a company for LESS than what it's worth.
Or buying *shares* of a company at a *price* that's *lower* than their *value*.
How much in *assets* does a company need to produce $1 worth of *earnings*?
#2 Thing To Look For In A Balance Sheet
Where do these assets come from?
In particular, how much of these assets has to be put up by the *owners* of the business -- as opposed to other sources of funding (like debt and float)?