3k? It is to laugh! Just how many billionaires are there in the world?

A fun survey of legible and illegible wealth

So if you look up the number of billionaires in the world, you’ll generally run across figures along the lines of 3,000. There’s ~3,000 billionaires, and the US has ~1k of those, for a healthy punching-above-our-weight of 33% of billionaires with ~3% of the world’s adult population.

But 3k? It is to laugh! I can find a few hundred more billionaires just by rummaging around in couch cushions! I can find 20% more just by looking beneath a single rock!

I maintain we have at least twice that many billionaires in the world!

So am I the crazy one, or is Forbes?

You decide, I’ll lay it out.

So Forbes, famously, calculates billionaire-hood by looking at primarily publicly traded companies, with a handful of the very largest and most prominent private companies, while excluding dispersed familial wealth.

But what more does Forbes leave out?

They leave out quite a bit, and this post is a fun dive into some lesser known wealth reservoirs that are essentially invisible to the Forbes methodology.

I’m not going to rigorously support all of these, just gesture at the space and do some simple math to convince you of “overlooked billionaire” potential - but I do only include things that are attested by numerous independent sources, so you should be able to do some independent research into any of these if you’re curious. I’ll also link the Google sheet with all my calculations at the bottom if anyone is curious.

So what do we have?

China

China is laughably wrong in Forbes, because it’s not transparent. BUT as anyone who has done business there can tell you, there are metric shitloads of USD billionaires in China, FAR more than the meager ~450 Forbes attributes. I mean, I personally know at least 3, and I’m nobody. I’m not sure what the billionaire base rate for nobody-to-billionaire knowing has to be, but 450 / 1.4B seems like a suspiciously low threshold.

Fortunately, we can do better than trying to estimate that.

Did you know China’s economy is bigger than the US economy, in PPP adjusted terms? This has been true for the past ~10 years!

And sure, we can’t use PPP to consider USD billionaires, but this is still relevant, because one thing China does have is greater wealth inequality than the legible areas Forbes looks at such as the USA and EU, as well as having the second largest economy in the world (clocking in around $18T to the US’s $30T).

Essentially, we can derive the number of USD billionaires for nearly any country with purely top level metrics like total GDP, population, Gini coefficients, and an income-to-wealth multiplier.

Why can we do this? Because income and wealth follows a power law everywhere in the world - there are basically two parts to an income or wealth distribution - a long (much greater than 50%) tail up to the median wealth/income per capita, and then a short tail of above-average wealth holders who pull the entire mean upwards significantly above the median ($50k median and $74k mean in the US, for example).

How long the poor long tail is, and how short the wealthy short tail is, is basically defined by Gini coefficients - how strong the wealth disparity is.

Because “billionaire” is a wealth rather than income statistic, we then need some multiplier to translate income into assets, and this varies too depending on asset classes and savings rates and things like that, but the multiplier doesn’t vary too much, generally between 3x - 6x, with a few lower ones. China, with a rigorous savings culture and being the manufacturing center for the entire world, has a 6.5x multiplier.

If you code all this up and crank a few countries and regions through it, you can get a top level Fermi-style estimate for any country you have GDP, population, and income Gini coefficients for. Then you can triangulate against wealth Gini coefficients and “total wealth” estimates for countries you have that data for, as a sanity check.

Back to China

So what does all this mathematical tomfoolery tell us about China? That China probably has closer to 1100 - 1300 USD billionaires versus the 450 we see in Forbes.

So sure, that’s just a MODEL!

But that’s fine, because we can triangulate a few ways.

For one thing, China’s Forbes billionaire rate is only 12% of Singapore’s, 18% of Hong Kong’s, 37% of South Korea’s, 64% of Taiwan’s, 86% of Japan’s, and 31% of the USA’s - all notably more legible, and all but the USA East Asian, so similar people, environments, and business cultures.

Another source of triangulation? The Hurun “rich list,” which does a much better job estimating Asian billionaires. They estimate 823 billionaires for China.

Any way we slice it, it’s pretty clear China is missing at least 300 - 800 billionaires from their tally, and there are similar arguments for South Korea and Japan.

I bring up South Korea and Japan, because the economic development model there literally revolved around massive family owned conglomerates - chaebols or keiretsu. As we’ll see here shortly, these types of companies are significant drivers in many countries’ economies.

But the Chinese underestimation doesn’t stop in China - it’s endemic throughout SE Asia, which brings us to our next area.

The Chinese diaspora that’s responsible for ~70% of economic activity across all of SE Asia.

This is another fun one that not many people know. Did you know the Chinese are a “market dominant minority” (MDM) in nearly all of SE Asia?

Amy Chua, of Battle Hymn of the Tiger Mother fame, the infamous how-to guide on “how to grind your kids into Harvard,” and whose family is MDM itself, actually wrote a book about this, World on Fire.

This was a fun one, because the book’s essential thesis was: democracy is a mistake for a lot of places where MDM’s drive the economy, because it just leads to authoritarian populists coming in with a platform of ‘those shifty minorities are hoarding all the wealth, let’s get them!’ Then they expropriate the wealth, inducing human and financial capital flight and wrecking their economies.

And indeed, so it has come to pass numerous times, which she goes over in the book.

Chua and the two daughters she successfully Tiger-mothered into Harvard

So, just like she was right about grinding her kids into Harvard (although really, if you start off with both parents as Yale Law professors, you kind of expected the kids to get into the Ivies), Chua is right about the vastly disproportionate wealth and influence of Market Dominant Minorities (MDM) in various countries throughout the world.

So, market dominant minorities are big players in SE Asia - and the “dominant” part of that descriptor is doing a LOT of work - we’re talking 66% - 80% of ALL private (as opposed to state/public) wealth, while being 1-4% of the population.

This is power laws at work! Essentially entire economies in SE Asia are being driven by Chinese ethnicity people with large private businesses. And you know what that means! There’s billionaires in them thar minorities!

And they are extremely illegible, because they’re often family-run conglomerates that are either fully private, or have some public companies and some private companies, and even the public companies have different levels of scrutiny and fewer public financial reporting requirements than developed-world public companies.

Examples of this are families like the Sy or Zobel de Ayala or Kuok families. And yes, I’m pointing to people that are Forbes listed - but my point is that they are the legible tip of an iceberg that’s responsible for literally ~70% of private wealth, and correspondingly large chunks of the GDP’s of entire countries.

Let’s just take it at a high level. Methodologically, we’re going to want to apply a power law even to that tiny fraction of the population, because wealth is even more Gini than income, and Pareto laws are fractally true - sure, the top 1% owns ~31% of the wealth in the USA, but did you know the top .1%, .01% and .001% owns 47/20/12.5% of that 31%? And the top .00001%, only 19 households, owns 5% of that?

An aside on inequality among the wealthy

One interesting thing that I found in my researching and modeling for this post was that the “wealthy” go through two inequality transitions - let’s consider the US:

The Gini coefficient, that infamous measure of wealth or income inequality, is measuring the discrepancy between a perfect 45-degree line representing pure equality, a simple linear line upwards as wealth increases smoothly and distributed equally across the population, versus the measured power-law effects of actual wealth / income distributions.

There’s some math here, but it boils down to the wealth Gini being ~0.85 for the entire US. This is against an income Gini of .40, which is far more equal. Basically for wealth in the US, the bottom third has negative wealth (net debt), another third or so is roughly neutral, and the top third has all the assets, for the wealth Gini of .85.

But one interesting finding - if you ONLY consider the top 1% by wealth (which has a floor at around $13-$15M in net assets), the wealth Gini is ~.5 - much more equal than the overall wealth Gini! This can be derived from the “percent of total wealth held” at each respective 1%, .1%, .01%, .00001% tier I outlined above.¹

Then when you get to “billionaire” territory, there’s another phase transition. Obviously the tranche of “billionaires” doesn’t merely have an average wealth of $1B, because there are deca and centa billionaires dragging the mean up. You can use the Forbes list to calculate the implied alpha, and thereby the implied Gini, given the USA figures - with a collective fortune of $6.8T across 902 measured billionaires, the mean billionaire wealth is $7.54B. The implied Gini from that is .76 - high again!

Back to SE Asia

So we know that a lot of wealth in SE Asia is illegible, and that a lot is owned by market dominant minorities. All this talk about Gini’s was as a basis for a model of a plausible number of billionaires given across SE Asia and other countries, because we need some plausible, data-based number to go off of when we’re considering potential gaps in legibility here.

What Forbes tells us is that there’s ~64 billionaires in SE Asia, which seems highly questionable. Just taking one country I know something about the wealthy people in, the Philippines, where 15 billionaires are labeled, we can detect a mismatch - these 15 are your Sy’s and Zobel de Ayalas and other owners of massive conglomerates, some of which are partially public, and so legible.

But overall, these 15 are responsible for some tiny fraction of employment (1%) and GDP (15%) and total wealth (3.7%) - if you add all their fortunes together, it amounts to something like 5% of just the wealth owned by the top 1% in the Philippines, and this is not what you’d expect - if we tie back to the best attested countries with solid data like the United States (14%), Germany (9%), UK (15%), Canada (17%), Australia (10%), and Sweden (27%), we see much higher numbers than that Philippines (5%), from putatively more equal places with much higher average incomes and economies.

So this suggests that at minimum, we’re underestimating just Philippines billionaires by probably a factor of 2-4, and that’s a conservative take.

We see a very similar deficit in Thailand, which has historically along with Brazil been one of the most unequal places in the world.

I’ll not bore you with the ins and outs, but broadly, the model I put together based on the known GDP, population, 1% wealth shares and measured Gini’s predicts about 2x more billionaires in SE Asia than we’ve measured, and this seems pretty safe and sane when triangulating by these other measures, too.

So we’re going to pick up another ~60 billionaires from here.

Other non-Chinese “market dominant minorities” in Latin America, Africa, etc)

So this MDM trend actually holds true for a a lot of the world, including Latin America, which is famously one of the most unequal regions in the world.

The MDM’s are non-Chinese there; they’re descendants of Spanish haciendero families or immigrant minorities from various places (Germany, Japan, the Levant), but they follow the general pattern of “huge chunks of GDP coming from their largely private family conglomerates.”

But across ALL of Latin America, with a cumulative GDP of $6.6T, we see only 94 billionaires on Forbes!

This shall not stand, sirrah!

You’re telling me in the most unequal region on earth, with ~$6.6T of GDP, which is ~1.5x bigger than Germany (171 billionaires), ~2x bigger than Italy (74), 3x bigger than Canada (76) only has 94 billionaires??

And you’ll notice I was not benchmarking on the extraordinary US there, I was benchmarking on famously equal and low Gini countries, all of which had a factor of 2-3x more billionaires weighted by GDP. Yeah, I don’t buy that - this is obviously a factor of legibility.

What does Latin America have that other places have less of??

The global cocaine market has an estimated value that was benchmarked at $100 - $150B annually in 1998. Just with a simple inflation adjuster, that would be $200 - $300B today. That’s an annual figure! This is an entire shadow economy that could be minting hundreds of fresh billionaires every year! And obviously with any relevant multiplier from flow to static wealth, this represents a multi-trillion dollar pool of value that we can power law some billionaires from - my modeling suggests there are conservatively at least another 200-400 billionaires here.

What else? Well, as we touched on earlier, there’s a general dynamic of privately held large conglomerates with fingers in everything - agriculture, forestry, banks, telecommunications and media, and construction.

There are massive privately owned companies in all these segments, and they are entirely illegible. The Odebrecht scandal gives a peek into the tip of the iceberg - an entirely private company that could bribe in hundred million dollar amounts ($350M in bribes just in Brazil), across ten countries - and this is just one company, there’s lots of these! We’ll touch on just how many there are in this space in the next section.

Overall, on the legible economy measures, my model predicts another ~200 billionaires from just the GDP’s, and probably another 300 from cocaine and other drugs. Let’s call it 500 total.

The ~90% of companies JUST IN THE US with >500 employees that are totally private

Coming back to the good old USA for a moment, and mirroring the MDM pattern - did you know that there is a vast mismatch between private and public companies, and ONLY ~10% of companies with more than 500 employees are public?

That means there are significant wealth reservoirs that aren’t accounted for by Forbes, which only considers ~200 private companies as source of wealth in total in the billionaire list.

What’s a good estimate of the >500 employee companies that aren’t considered by Forbes, but still generate enough wealth to make some shareholders billionaires?

Broadly, there are at least 19,000 privately held companies in the US with at least $100M or more in revenue. That is just a mind-boggling figure to me.

But if we crank them out with average values based on 10-20% nets and 5 year multipliers averaged across all industries, we get around 450 billionaires that are probably living there in illegible “private company driven wealth” land.

However, that’s duplicative of the legible ~200 Forbes ones, so there’s probably only ~250 extra American billionaires from private companies. What did my model predict based on America’s economic fundamentals? About 200.

But more importantly, this "hidden reservoir of private company wealth” is broadly true all across the world, and is the dominant norm in the great majority of countries. Sure, there’s certainly a power law discount in terms of minting literal billionaires, but the fact that Forbes doesn’t even look at owners of 90%+ of companies in the US(!) tells you a lot.

This should indicate, if nothing else, how conservative I’m being here.

But you know what else Forbes doesn’t consider?

The Middle East and North Africa

Did you know Forbes says there’s only 15 billionaires in all of Saudi Arabia?

Did you also know the House of Saud (15k total, around 2k core members) holds at least $1.5T in wealth, while the entire GDP of Saudi Arabia is ~$1.2T? I mean, if you just divide that across the 2k core member group, you get a mean wealth of $750M per House member, and at least 200 additional billionaires by a fully-unequal power law estimation (when you would expect families to have lower Gini indexes internally than economies full of strangers) - just in Saudi Arabia!

The howlers continue with UAE (6), Qatar (2), Oman (2), Dubai (0), and North Africa (10).

Just hitting some high points, but the Al Nahayan family (~200 members) in UAE is worth >$300B, and that’s before you consider the >$1T in sovereign wealth funds they direct.

Dubai, which doesn’t even appear on the Forbes list, has the Al Maktoum’s (low hundreds), with family patriarch Sheikh Mohommad alone worth over $18B, the rest of the family worth in the tens to hundreds, and with another $400B in a sovereign wealth fund the family controls.

The House of Thani (3-5k total, 3-5 core) in Qatar is worth ~$300B or so, and controls $500B in the wealth fund, with the Emir’s personal worth outside of the family $2-3B.

Aayatollah Khamenei controls a $95B annual revenue business conglomerate in Iran, also not on the list, with 20-30 core high level decision makers.

Just on the raw GDP figures, setting the House of Saud and other leadership families totally aside, we’d expect a little over 200 billionaires from the MENA region, and the listed Forbes ones are only 72.

Throw in the families, and I’m estimating there’s at least 1k more billionaires in this region.

A fun aside on Europe and taxes and equality

Okay, so I’m talking about the model I put together here. One interesting thing about it - it predicts way more billionaires for Europe than actually exist in Forbes. Like 600 - 700 more.

I’m not even going to fight for including them here, either!

But I think that if it’s true, and Europe has ~650 fewer billionaires due to more equality and higher taxes, that this is A) interesting, and B) worth looking at a few examples.

When I was QC-ing my model’s results by country and deciding if it made sense or not, I ran into some interesting mismatches. For one thing, Luxembourg has zero Forbes billionaires despite very high average incomes and wealth levels, while tiny Monaco, roughly 20x smaller, has at least 2 - my model was predicting 40 billionaires in Luxembourg by the numbers!

What’s the mismatch? Monaco has no personal income tax, and Luxembourg has a ~40% top bracket income tax. This can drive out many tens of prospective billionaires that could otherwise be generated or persuaded to live there, apparently!

Another data point along these lines? Sweden, famously egalitarian and as Scandinavian as they come, punches way above it’s weight on the billionaire front! It has 13 - 20 predicted by the economic numbers, but actually has 45 Forbes listeds! That’s a huge differential!

What’s the answer? Sweden has pursued a tax policy that allows billionaires to stay and thrive - they eliminated the inheritance and gift tax in 2004 and the wealth tax in 2007. If you build a $10B company, you can pass that to your heirs tax free - in the US, that would be taxed at 40%. And indeed, you see some names you’d recognize in the Swedish billionaires, who derive their fortunes from H&M, TetraPak, Spotify, and Mojang / Minecraft, among others.

Political wealth (dictators and kleptocrats, oh my)

Another area Forbes specifically eschews in its tally - political wealth. We’re talking corruption here. Corruption of the billion dollar palace variety, of the “flying an additional private jet just for your luggage” flavor, of the buying Manhattan skyscrapers on a lark type. You know what I’m talking about!

Forbes doesn’t count these, because they’re both relatively illegible AND controversial for realpolitik reasons.

But we can sure count them!

It actually doesn’t require all that much GDP to mint corruption billionaires.

• The Philippines did it with the Marcoses at ~$200 - $600 GDP per capita, and $6 - $30B total GDP (from beginning to end of term).

• Indonesia did it with Sukarno ($100 - $150 and $9-$15B GDP) and Suharto ($150 - $530 and $16B - $110B).

• Zaire’s Mobutu Sese Seko did it at $145 and $2.5B and Angola’s Jose Eduardo dos Santos did it at $440 and $3.3B.

Surprisingly, you basically need ZERO money in an economy to become a kleptocratic billionaire! One or two hundred bucks per person will do the job.

Moreover, it’s relatively systemic - if one person / family can become corruption billionaires by reaching one of these offices for a given term duration, so can others. As just one example, during the single term Dmitri Medvedev was president in Russia, with Putin Prime Minister, he accumulated ~$1.2B in bribes.

Most of these examples were literally the bottom of the corruption index for the ones that were around recently enough for the corruption index to exist - but Angola was 142 /163, or bottom 12%. I’ll consider any country eligible here that’s bottom 15%, because the actual GDP doesn’t seem to matter at all, and that’s where Russia appears in the list, so it seems pretty apropos.

Then lets estimate that every 10 years on average, they get overthrown or kicked out and mint another corruption billionaire over the succeeding 10 years, and that this has been going on since the 70’s - how many do we have total? 50 years * 27 most corrupt countries = ~250 more billionaires just from corruption worldwide! Let’s call it 200 to be conservative.

Hidden and tax advantaged assets

This is another fun one that might be relatively lesser known.

Most people have some vague recollection of Swiss bank accounts as being a place where people from all over the world hid their ill-gotten wealth.

This hasn’t been true since 2010, when the Swiss became narcs, but the general idea is true - there are definitely large wealth reserves in various deliberately illegible financial reservoirs.

Some fun ones I know about just off the top of my head:

Trusts and foundations - I’m talking self-settled South Dakota trusts, their Montana, Nevada, and Wyoming competitors, private 501c3 family foundations, and more. Completely legal entities and instruments that are custom built to protect assets and make it illegible and / or tax-advantaged.

Just South Dakota trusts are estimated to have $1T in assets alone. If you do some reasonable discounts for the lesser known states and add the known private foundation assets, you reach a conservative ~$3T private illegible wealth floor figure, and we know that ~$3T is divided amongst 8-20k accounts conservatively (for example, 81 accounts at the one leaked SD trust company, and there’s ~108 of those overall, and if each has ~100 accounts, that’s roughly 10-11k total accounts), for ~200-350 more prospective billionaires. There’s probably a lot of overlap with the legible known billionaires out there - but we’ll get to that a little later.

In the HSBC Swiss bank account leak in 2015, 100k clients had assets >$180B at just one Swiss bank. This gives us some useful triangulation figures to estimate from, and we can infer that for every ~4k Trust accounts, or every ~25k Swiss accounts, there will be one billionaire on average - but did you notice that just one Swiss bank leak had 100k accounts? There were 3.1M Swiss bank accounts total at the time, for another 124 billionaires.

Notably, you do NOT have to be a US citizen to open a South Dakota trust - so there’s actually a good number of overseas settlors.

Cook islands accounts - known more for their obdurance in the face of legal requests, Cook Islands accounts are the go-to instruments for rich people in legal trouble who want to protect their wealth from judgments. I’m not even going to bother here - this is clearly the range of mere millionaires, most billionaires don’t need to stoop to this, they just hire sufficiently vicious attack lawyers to dissuade you.

Crypto wealth - Only considering Bitcoin, at current prices there is ~$150B+ in dark wallets like Satoshi’s that are attributable single wallets large enough to be billionaires without active transactions, and $22B was seized from Silk Road alone, indicating that many of the other darknet drug market owners are likely to have been able to attain billionaire status (and there’s been a lot of those). This adds up to around 7 incremental, taking out the known 6 crypto-holdings billionaires and 19 known crypto platform billionaires. Then if you consider the value of all other crypto, it’s another ~$1.3T in assets, across an estimated 400M altcoin holders worldwide. If you do the power law math on that distribution, you’ve got another 5 or so additional billionaires. This was basically rounding error, which kind of surprised me.

Other hidden assets - there’s going to be a lot of overlap here with the 90% of private companies, trusts, foundations, and other things here, but many people hold wealth like planes, buildings, and real estate in LLC’s, holding companies, and other legal entities often with deliberate obfuscation (webs of LLC’s owning other LLC’s, SPV’s, fiduciaries, funding and legal ownership coming from lawyers’ pooled client / IOLTA accounts on behalf of undisclosed clients, and so on). Known IOLTA assets are on the order of $100 - $200B per year. That’s not a ton, and it’s not limited to American citizens (anyone can hire a US lawyer), but it’s probably room for a few more hidden gigabuckaroos here and there.

If only there was some sort of “catch and release” for billionaires, so we could estimate the hidden versus legible proportions…

Amazingly, there is a billionaire “catch and release”

Leaks like the Panama Papers, Pandora Papers, Paradise Papers, and the HSBC Swiss bank account leak have fingered Forbes billionaires a couple of times.

The Pandora Papers, specifically those regarding South Dakota trusts, is our best example. 130 Forbes billionaires were found in them, with 81 accounts having ~$100B in assets.

So there’s 130 leaked for the SD trust Pandora Papers leak - but remember our triangulation math showed that there’s likely 200-350 billionaires using them?

Now I can make some catch and release estimates using the Lincoln index equation, using 3k as the marked pool first draw (E1), 200 - 350 for the modeled independent second draw (E2), and 130 as the recaptured population (S), we get a lower bound of 4.6k true billionaires, and an upper bound of 8k.

Summa:

My best estimate on the number of billionaires:

Okay, what do we have here? So just based on GDP, we have about 2.3k “extra” predicted billionaires, and this is China, other countries, the 90% of private companies in the US and elsewhere, and things like that.

But then on top of that excess, we have some known incremental sources of billionaires that aren’t measured for whatever reasons, and these include drugs, the Middle East, crypto, kleptocrats, and illegible wealth in trusts and foundations.

At the end of the day, I think I’m at around 6-8k, or 7k total billionaires in the world - this was actually mildly disappointed to me, because I had been Fermi estimating in my head that it would be around 10k before I kicked off the data and modeling pulls for this post. Still, I think I was conservative enough in all this that I could scrounge up another 3k-5 if I really needed to, which would require deep dives into Pandora and Panama Papers and things like that - staying at a pretty safe floor seems fine relative to that amount of incremental work.

Also, America takes a big downgrade - from punching 10x above population weight at ~30% of the world’s billionaire total, down to around 15%.

Also, “American exceptionalism” goes out the window another way - far from being the bastion and light of capitalism and human capital leading the world, “richer in PPP terms” and nominatively communist China goes up to 93% of our billionaire rate if I’m right, and that’s with a MUCH higher baseline population. Harsh.

But then again, it also fits pretty well with the one solid “catch and release” sample we got!

So, mission accomplished?

Why not.

If you’re interested in the Google sheet with the data and calcs, it’s here.

1. Specifically the tiers as used and pasted from my jupyter notebook:
(15_800_000, 0.01), # Top 1% (x/x_99 = 1.0)

(73_509_000, 0.001), # Top 0.1% (x/x_99 = 4.29)

(304_336_000, 0.0001), # Top 0.01% (x/x_99 = 18.2)

(1_600_000_000, .00001), #3417 a third of which are billionaires, pulling the average up

(7_100_000_000, 900 / total_us_adult_pop), #Billionaires average more than $1B

(16_400_000_000, .000001), # top 350 families

(85_000_000_000, 19 / total_us_adult_pop), # top 19 families