Harry Browne’s Permanent Portfolio and Me

Harry Browne was a man who lived mostly in the last century. He was an author, investment adviser, newspaper syndicator and Libertarian. He ran twice for president on the libertarian ticket in 1996 and 2000. He set up a portfolio for “all conditions” His portfolio consisted of 4 assets stocks bonds cash and gold in equal amounts.

If you Monte Carlo this you can withdraw 4% for 30 years with 97.55% success

If you have 3 bad years of SORR up front (the worst stress) you survive 94% of the time.

For reference a BH3 survives 3 yr bad SORR for 30 yr at 4% withdrawal 38% of the time.

So Harry Browne’s Portfolio is pretty damn safe. It does just what it says it will do. The Harry Browne does’t hit a home run, no 9M value at the 90% level, but it gives you reliable 4% come rain or shine. The HB looks like this on the EF plane.

It is not on the EF so lets adjust the AA and place it on the EF

Lets Monte Carlo that AA

With 3 years of bad SORR and 4% inflation adjusted withdrawal you survive 99.92% of the time and have half a mil left at 30 years in the worst case


This post is a radical demonstration of the power of diversity using the EF as a guide. If your goal is set it and forget it 4%/yr year in year out the EFPP is the ticket (efficient frontier permanent portfolio). The reason it works is a relatively high rate of return (7%) and a low SD (5%) . The BH3 (7%return 12% risk) fails 80% of the time at 5 yr bad SORR. The trick to the EFPP is don’t change anything. If the market is up one year do not be tempted to move some cash into stocks.

BTC badly distorts the EF

because of it’s out sized return and vol over the last 8 years but you can still do the calculation

This is my portfolio simplified around 58% VTI as the stock, 9% BND as the bond 7% EDV as the bond (total 16%), 16% as the cash in a high yield represented as BIL and 7% GLD. As you can see the portfolio sits on the EF. I can’t Monte Carlo this portfolio because the MC needs 10 years of data to run but I’m willing to give this a try. Disclaimer: I do not recommend this portfolio. If you do it don’t blame me. I think it’s constructed on sound Modern Portfolio Theory principles of maximal non correlated diversity. Do I expect 16% return? No. What I expect is smaller loss in a more robust portfolio, similar to the EFPP, if there is Armageddon. It amounts to that I’m substituting a source of high volatility and high return that is transparent and not open to market manipulation and completely non correlated to stocks for a portion of the vol and return normally allocated to stocks. The BTC bet is small only 3%. Notice the ratio of return to risk is like (not the same number but the same direction) as the risk and vol in the EFPP. with a greater than 1 Shape’s ratio. If you look at the BH3 risk is 12 return is 7. EFPP risk is 5 return is 7, less than 1 Sharpe’s In this portfolio risk is 10 return almost 17, so this portfolio and EFPP are “the same” Sharpe’s type vs BH3 which is different.

There has been a ton of analysis done on Harry Browne PP plus a lot of biased commentary. It’s worth studying but comparisons are often not logical but replete with bias discounting HB for whatever the author is hawking.

9 Replies to “Harry Browne’s Permanent Portfolio and Me”

  1. Ahh the memories. Back in the day I was in 100% equities, almost all individual stocks. My 2 exceptions were BRKB and the Permanent Portfolio. It performed as advertised and was especially stable during downward corrections. Both were the anti-stupid part of my portfolio, just in case too many of my stock picks turned out to be fools gold. Nice enhancing modification putting the PP on the EF.

  2. Do you think almost 60% stock is too risky if, as some of the videos you posted previously stated, we are heading for 1937 and a decrease, potentially permanent, in equity valuations?

    1. I’m about 57% equities. What matters is what else you own, if you will still have a job etc. Imagine your portfolio as a series of buckets each contains a different asset label them BIL BND VTI GLD and BTC. Your net worth is the sum of the buckets. Lets say you have a BUNCH of VTI and economic conditions come along and remove half of your VTI in the bucket how would you protect yourself? You would take some from VTI and distribute it into the other buckets before the economy came and snatched it away. You wouldn’t want to sell all the VTI because it continues to make money for you, just some. Since the networth is the sum of the buckets you still have the same networth. If you want to keep the same level of risk you can do that with a little BTC. Basically BTC is risky but it does no correlate with stocks which are also risky so you substitute one risk for another when you buy BTC but when the economy snatches the VTI the BTC stays unaffected directionally. It still has its risk but it is independent of stock risk. This is the same with all of those assets each has its own independent risk not correlated to stocks so when stocks crash the other buckets maintain their risks separately. While I reduced my stock risk, by buying BTC I kept my overall risk about the same, and actually improved my return profile so if stocks crash I could either loose a little, stay the same, or even make a little depending on how the other buckets do and the relative ratios of each. The goal is to at most loose a little while retaining a similar risk profile.

  3. Solid pick for the risk averse investor – I have a few friends that come to mind who could likely stomach necessary investing better with a Harry Browne approach. Appreciate the analysis.

    1. An efficient frontier optimized HB portfolio is actually a pretty awesome portfolio. 34% VTI, 59% BND and 7% GLD would give 6.15% return and 6.06% risk. You can put some of that BND in cash in a high yield and not loose much if anything since that would tend to reduce the risk some as well. VTI has a return of 9% but a risk of 15%. At 4% WR a 34/59/7 has a 97% survival for 30 years (1st case failure at 19 years) compared to a BH3 which has 90% survival and 1st case failure at 10 years. I would turn a percent of the BND to cash in a high yield or CD ladder.

  4. Hi Gasem ,

    Can you please tell me what software or website you used to conduct the Monte Carlo scenarios in this blog post?

    I believe for plotting Efficient Frontier , you used

    What about Monte Carlo scenarios?

    May I ask what software you used for that?

    Any other software you like using for creating DIY analysis and graphs and charts in your blog posts?

    Please advise.


Leave a Reply

Your email address will not be published.