Does the 2012 JOBS Act make historical market simulations unreliable?

[Ex Pat]

Before 2012: once a company had 500 shareholders, it was basically forced to go public and disclose everything. So fast-growing companies IPO'd early — Microsoft in 1986, Google in 2004 — and ordinary investors could buy in and ride the growth.

Even employee stock was a forcing function: option plans put shares in thousands of employees' hands, pushing companies over the 500-holder line — one of the classic reasons fast-growing companies had to IPO.

The 2012 law changed three things:

1. Employee shares no longer count.
2. The registration threshold was raised from 500 to 2,000 shareholders — and only 500 of those can be non-accredited (ordinary retail) investors, so a company can now have 2,000 wealthy shareholders and stay private.
3. It became much easier to raise private money — companies can openly advertise private funding rounds to rich investors.

The result: companies like Stripe and SpaceX could now raise billions without ever listing on the stock market. All that growth used to flow to public index investors; now it stays with private investors.

I understand that Vanguard published a 2025 report saying companies still go public at the same average age as before.
But companies like Stripe raised billions without ever listing, and SpaceX only went public in 2026 — after 24 years of growth, at a $1.77 trillion valuation.

Previous game-changer companies like Apple and Microsoft and Google and Amazon became public much earlier in their growth stage.

And index returns have always been driven by a handful of winners — that's exactly why this matters. Those few companies that make or break S&P 500 returns are now precisely the ones being kept out of the index during their steepest growth.

My question: given that the biggest growth stories no longer reach the public market until they're already mature, are simulations based on past US market returns overstating what index investors should expect going forward? By simulations I mean the standard Monte Carlo FIRE projections that assume US index returns of the last 50–100 years repeat.

where I first encountered this :

 

[Baptiste Wicht]

My question: given that the biggest growth stories no longer reach the public market until they're already mature, are simulations based on past US market returns overstating what index investors should expect going forward? By simulations I mean the standard Monte Carlo FIRE projections that assume US index returns of the last 50–100 years repeat.

There may be a slight impact. But I don't think that a change over 13 years would really impact historical simulations over 150 years.

Additionally, companies like OpenAI, Anthropic, and SpaceX are all AI-valued, and we are in an AI bubble. So I do not think they are really representative of what we want in the market.

It probably impacts our ability to use historical simulations to "predict" the future, but this tool was always imprecise. That is why we should build some margin of safety. And this remains in my opinion the best tool we have for planning FIRE.

 

[Ex Pat]

are all AI-valued, and we are in an AI bubble

Indeed. But the mechanism shows up in plenty of non-AI companies that grew on private capital without going public:

• ByteDance (2012) — ~$480B — media (TikTok)
• Stripe (2010) — ~$159B — payments
• Checkout.com (2012) — ~$40B — payments
• Ant Group / Alipay (2014) — ~$87B — fintech (IPO cancelled in 2020)
• Revolut (2015) — ~$75B — banking/fintech
• Ramp (2019) — ~$60B — fintech
• Plaid (2013) — ~$14B — fintech
• Binance (2017) — ~$70B — crypto exchange
• Shein (2008) — ~$67B — e-commerce
• Canva (2013) — ~$32B — software/design
• Discord (2015) — ~$15B — social/gaming
• Waymo (2016 spinout) — ~$126B — autonomous driving
• Grammarly (2009) — ~$13B — software
• Anduril (2017) — ~$90B — defense

Anyhow, I'm not claiming that no companies go public anymore.

Putting things differently:

• Microsoft had ±1.2K employees when it went public.
• Amazon : 256 employees.
• Google : ±2K employees.
• Apple : ± 1K employees.
• Facebook : ±3K employees.

All these companies went public early in their growth, and the 500-shareholder rule was one of the main forcing functions — employee option plans alone pushed companies over the line.

The Chinese companies didn't stay private because of the JOBS act, but still those are large monoliths that grew a lot without giving access to retail shareholders during their global expansion.

All I'm saying is that Monte Carlo projections are calibrated on a world where companies went public early and retail investors could join their growth. That doesn't mean the market won't grow, or that nothing goes public anymore. It just means that the flagships most likely to drive the next few decades of index returns spent most of their expansion outside the index — so expected returns are likely somewhat lower than the historical sample implies.

And I might also be wrong.

 

[Ex Pat]

It probably impacts our ability to use historical simulations to "predict" the future, but this tool was always imprecise. That is why we should build some margin of safety. And this remains in my opinion the best tool we have for planning FIRE.

ah, you actually said the same thing

 

[Baptiste Wicht]

All I'm saying is that Monte Carlo projections are calibrated on a world where companies went public early and retail investors could join their growth. That doesn't mean the market won't grow, or that nothing goes public anymore. It just means that the flagships most likely to drive the next few decades of index returns spent most of their expansion outside the index — so expected returns are likely somewhat lower than the historical sample implies.

I agree on that indeed.

Any historical simulations (backtesting, Monte Carlo, and bootstrapping) are limited in that we try to predict the future even though we can't.

If you are planning for lower returns, how do you plan for that in your FIRE Plan?

 

[HerbertBTroupe]

Before 2012: once a company had 500 shareholders, it was basically forced to go public and disclose everything. So fast-growing companies IPO'd early — Microsoft in 1986, Google in 2004 — and ordinary investors could buy in and ride the growth.

Even employee stock was a forcing function: option plans put shares in thousands of employees' hands, pushing companies over the 500-holder line — one of the classic reasons fast-growing companies had to IPO.

The 2012 law changed three things:

1. Employee shares no longer count.
2. The registration threshold was raised from 500 to 2,000 shareholders — and only 500 of those can be non-accredited (ordinary retail) investors, so a company can now have 2,000 wealthy shareholders and stay private.
3. It became much easier to raise private money — companies can openly advertise private funding rounds to rich investors.

The result: companies like Stripe and SpaceX could now raise billions without ever listing on the stock market. All that growth used to flow to public index investors; now it stays with private investors.

I understand that Vanguard published a 2025 report saying companies still go public at the same average age as before.
But companies like Stripe raised billions without ever listing, and SpaceX only went public in 2026 — after 24 years of growth, at a $1.77 trillion valuation.

Previous game-changer companies like Apple and Microsoft and Google and Amazon became public much earlier in their growth stage.

And index returns have always been driven by a handful of winners — that's exactly why this matters. Those few companies that make or break S&P 500 returns are now precisely the ones being kept out of the index during their steepest growth.

My question: given that the biggest growth stories no longer reach the public market until they're already mature, are simulations based on past US market returns overstating what index investors should expect going forward? By simulations I mean the standard Monte Carlo FIRE projections that assume US index returns of the last 50–100 years repeat.

where I first encountered this :

That's a fascinating point about private markets capturing early mega-growth. However, since a 13-year policy shift has a minimal footprint on a 100-year historical dataset, and tech giants are often heavily hyped, standard Monte Carlo simulations remain our best planning tool—provided we maintain a healthy margin of safety

 

[Ex Pat]

However, since a 13-year policy shift has a minimal footprint on a 100-year historical dataset,

Exactly, the policy shift effects could only be seen in the next decades.

If you are planning for lower returns, how do you plan for that in your FIRE Plan?

I don't know TBH. The reason why I look into investments was diversification of wealth and avoiding the inflation cost. The FIRE community provides advice for new investors. So I kinda mechanically assigned a FIRE goal to my IPS (out of thin air, without doing simulations), but I didn't truly make a plan yet.

This year I've learned a lot about the financial markets, I need a break from this topic now. Next year maybe I will start doing simulations and decide whether to keep the investment portfolio only as a 4th pillar and diversification, or really make plans for FIRE.