Summary I construct and examine three ETF-based portfolios that follow Harry Browne’s “Permanent Portfolio” investment advice. I compare the best of my three portfolios with a mutual fund and an ETF that are also based on Browne’s advice. I compare my ETF retirement portfolio against the Permanent Portfolio candidates in terms of both total return and dividend yield. My last article drew some rather critical comments, some of which were quite interesting; 1 one comment in particular brought up the Harry Browne Permanent Portfolio – something I had not encountered before. 2 In this article I take a look at the Harry Browne Permanent Portfolio (hereafter, HBPP ), what it might look like populated with ETFs, and how it might perform. The Strategy Harry Browne’s strategy is articulated in his book, Fail-Safe Investing: Lifelong Financial Security in 30 Minutes . It also seems to be a regular topic of conversation on the internet. 3 The strategy is based on a simple premise: a safe portfolio will be diversified according to types of investments that provide performance during particular economic environments. Browne identified four types of holdings and the economic state each type was geared to: U.S. stocks , which will perform strongly during a prosperous economy (apparently using an index based on the S&P 500); Long-term Treasury bonds , which will do well in prosperous times, but also during deflation; Cash (money-market funds) in recessions; Precious metals (specifically gold) during inflationary periods. 4 According to Browne, the portfolio should be divided equally between the four sections: Browne felt that index funds would be appropriate for the stock portion, and preferred a fund based on the S&P 500. He also preferred that the precious metal be gold, in the form of gold bullion coins . Other than those preferences, nothing seems to have been written in stone – hence this is more of a portfolio allocation strategy , rather than a hard and fast portfolio description. HBPP is intended to be permanent ; not that the investor should expect to be able to set it up and walk away, but that it should only require minimum maintenance. Periodic rebalancing is desirable, either annually or – as suggested by many – when any section of the portfolio becomes more than 35% or less than 15% of the portfolio overall. Certainly, if any holding in the portfolio becomes unsuitable, it should be replaced. A Suggested Portfolio ETFs became popular about 25 years after Browne developed his strategy, but it seems that ETFs would be particularly well suited to HBPP ; 5 all that is needed is to choose a fund for each section of the portfolio. Here’s where we encounter some “controversy.” Browne suggests that one should use at least three funds to cover stocks, 6 but there is indication that Browne himself used only one. I tried three formulations. In one (the first one I tried) I used: SPDR S&P 500 ETF (NYSEARCA: SPY ) SPDR Gold ETF (NYSEARCA: GLD ) iShares 20+ Year Treasury Bond ETF (NYSEARCA: TLT ) iShares 1-3 Year Treasury Bond ETF (NYSEARCA: SHY ) / SPDR Barclays 1-3 Month T-Bill ETF (NYSEARCA: BIL ) 7 I will refer to the above portfolio as HBPP1 . The second portfolio I considered used three funds for stocks and two for precious metals: Guggenheim S&P 500 Equal Weight ETF (NYSEARCA: RSP ) First Trust NASDAQ-100 Equal Weighted ETF (NASDAQ: QQEW ) SPDR Global Dow ETF (NYSEARCA: DGT ) GLD ETFS Physical Precious Metal Basket Shares (NYSEARCA: GLTR ) TLT iShares Short Treasury Bond ETF (NYSEARCA: SHV ) I will refer to this portfolio as HBPP2 ; for the stocks, RSP receives 12.5%, with QQEW and DGT getting 6.25% each. 8 GLD receives 25% until 22 October 2010, when GLTR was issued; at that point GLD receives 15% of assets and GLTR receives 10%. The third portfolio represents an effort to simplify HBPP2 by returning to the one-ETF-per-section mode: Oddly enough, I called this formulation HBPP3 . I tracked the performance of each of these portfolios back to 1 January 2007, and ran a test through 23 October 2015 for each. I decided to “compromise” between the two rebalancing schemes suggested by Browne: I rebalanced HBPP1 in June 2007 to switch from SHY to BIL, and switched HBPP2 in October 2010 to add GLTR. Other than that, I checked distribution of assets in June and December of each year; if a fund was weighted more than 30% of the portfolio as a whole – or less than 20% of the portfolio – I rebalanced. The following chart shows how the three portfolios performed: (click to enlarge) HBPP1 (which was the first formulation I tested) turns out to lose quite decidedly to the other two formulations. I initially tried HBPP2 to see if there was a difference between using a single equity fund and the three-fund equity position. 9 I also added GLTR to broaden the precious-metal holdings a little. HBPP2 clearly defeated HBPP1 , and it seems it does so shortly after the rebalancing that was needed on 2 January 2009 – after that point HBPP1 started to gradually lose traction. All else being “equal,” it seemed the combination of three equity funds and two precious metal funds did the trick. 10 What interested me the most at this point was the extent to which swapping RSP for SPY made any difference, which was why I tried HBPP3 . And clearly, HBPP3 outperformed HBPP2 , and there is some reason to believe GLTR dragged HBPP2 down. GLTR was added to the portfolio on 22 October 2010 (at which time the portfolio was rebalanced), and it is after that point that HBPP2 begins to fall off. 11 There would not be another rebalancing until 1 July 2012(and then, for HBPP3 only – HBPP2 would not get rebalanced again until 1 July 2013), and by that time HBPP2 had noticeably diverged from HBPP3 . 12 So, HBPP3 , with only one ETF for each of the four sections of the portfolio, is the one which provides the best performance of the three portfolios tested. The following chart shows the performance characteristics for each of the four component ETFs: (click to enlarge) The performance for the components gives a good indication of why the performance of the permanent portfolio is fairly smooth and constant. As one component decreases the others increase, and the general trend for all of the funds is to at least stay constant. Most elements move fairly regularly, the only exception being GLD. The precious-metal fund has a dramatic rise from late 2008 into late 2011, then begins to cool off. The drop in gold prices has been persistent for the past three years, although it is expected to become somewhat bullish as early as 2016. 13 That attitude is not shared by everyone, however. 14 The only ETF that seems completely dependable is SHV, which is up just over 1% for the 8-year period tracked. Overall performance The following chart shows the overall return from HBPP3 since 2007: (click to enlarge) The portfolio earned $1,581.79 in dividends and interest over the nearly nine-year period tested, resulting in a total return on the initial $10,000.00 investment of $7,005.32, or a total return of 70.05%. This amounts to a compound annual growth rate (CAGR) of more than 6%. Its yield (TTM) is 1.80%. How HBPP Compares to “Browne’s Own” Permanent Portfolio In 1982 The Permanent Portfolio Family of Funds offered a mutual fund, Permanent Portfolio (MUTF: PRPFX ). Harry Browne was himself one of the founders of the fund; however, PRPFX is not strictly structured as Harry Browne had specified in his writings. Still, it strives to achieve the same sort of equilibrium between portfolio components; specifically: 35% Dollar Assets (government & corporate bonds) 25% Precious Metals (20% gold, 5% silver) 10% Swiss Franc Assets 15% Natural Resource & Real Estate Stocks 15% Aggressive Growth Stocks The fund’s performance is represented below: (click to enlarge) The Global X Permanent ETF (NYSEARCA: PERM ) tries to be faithful to the spirit of Browne’s writings, but still does not quite capture the essence of the four-way division of the portfolio: 57. 92% U.S. Bonds, Financials and “Other” ( 48.91% short-term bonds!) 13.26% Precious Metals (9.71% gold, 3.55% silver) 28.83% Equities (short- and large-cap & international equities, real estate & materials) PERM ‘s performance since its inception in 2012 is reflected below: (click to enlarge) PERM ‘s performance to date seems to indicate why ETF.com judges that the fund “lacks investor interest”, and why they “see high fund closure risk.” 15 A Comparison A comparison of the three representations of Harry Browne’s Permanent Portfolio would seem appropriate. Furthermore, since the discussion of Browne arose from my article on changes to the ETF retirement portfolio, I thought it would be interesting to see how the retirement portfolio stacked up. In the following chart I provide the performance and total returns for HBPP3 (I will simplify this to HBPP ), PRPFX and PERM , along with the chart for ETF/R-A . 16 The comparison runs from 1 January 2014 through 15 October 2015; all performance data pertains to that period of time only. Each portfolio began with an initial cost basis of $10,000.00. Dividends/interest were not reinvested. (click to enlarge) PRPFX clearly has not fared well over the past 20 months. Even though it is up more than 243% historically, it has seen a drop of 1.81% over the trial period. 17 On the other hand, PERM – which had been losing – actually experienced a 5.11% growth over this period. Its growth was not overly facilitated by its annual dividend, which was $0.23/share, or $101.16 for purposes of the test. 18 Both HBPP and ETF/R-A put up superior numbers, with the retirement portfolio showing a 7.82% total return, and my rendition of the permanent portfolio topping that at 9.74%. Should I switch my money to HBPP ? Good question. Assessment The answer: no – at least, not yet . The following graphic gives a clearer (at least, less cluttered ) picture of HBPP and ETF/R-A : (click to enlarge) HBPP does outperform ETF/R-A by 192bps – nearly two full percentage points – and that is significant. However, scoping the full chart, it is clear that ETF/R-A is not exactly a performance slouch. Indeed, it seems to be only during the past three or four months that HBPP has led it consistently – otherwise, it seems that ETF/R-A has the advantage. A possible reason for the retirement portfolio’s recent performance is that it is equity-heavy compared to HBPP , which has “non-performing” short-term bonds and reasonably performing long-term bonds . Could these have countered any equity-based downwards pressure? Well, that is why Browne thought a permanent portfolio should have them. The real question is: will HBPP continue to outperform the retirement portfolio? Until I get a clearer picture over the next several months, I will stay with the portfolio I have. However, HBPP seems to have a lot going for it over the long haul. It is worth keeping an eye on. A Word About Yield Some readers of my last article noted – with some displeasure – the appearance that some changes made to ETF/R-A reflected a preoccupation with yield, and the chart above illustrates another reason why I am reluctant to switch from ETF/R-A to HBPP : the retirement portfolio beats the yield of HBPP by a very healthy 417bps . So, why a focus on yield ? From the outset last year, one of the ideas behind the retirement portfolio has been achieving a yield of at least 5% . This is in response to the traditional wisdom that advises retirees to figure on withdrawing about 4% from their portfolios per year. It struck me at the time – and still does – that it is possible to secure more than that in yield while still providing for a modicum of value growth. If the retirement portfolio continues to perform along its current lines, I will have achieved my goal . Why focus on earning the money in dividends, rather than capital gains? There are two reasons: First , making 5% or more in yield provides the ready cash to withdraw from the portfolio without disturbing my capital base. Otherwise, if I need funds I have to sell shares. Selling shares (if I am at all successful in choosing good ETFs) would involve capital gains (in addition to diminishing my capital). Unless the account is tax-advantaged, I’m going to have to pay higher taxes on those gains than I would on an equal amount of dividends (supposing the dividends were all eligible for the lower tax rate). If accumulated dividends become greater than foreseeable needs, I can always re-invest the excess. And since one has to pay taxes on dividends regardless, using them doesn’t constitute an extra tax burden. 19 Second , selling shares may involve transaction costs – which costs are typically due even if one has the shares in a tax-advantaged account. Investor-managed accounts may not involve large costs (and some accounts are free – at least, to a point), but transactions performed on a regular basis are going to add up. Dividends, on the other hand, do not usually incur transaction costs when withdrawn – it is simply a matter of transferring funds. Some institutions may even offer the service of automatically paying one’s dividends to one’s checking account instead of leaving them in the investment account. This would typically be done without charge. 20 Ultimately, this becomes a matter of personal taste and investment style. My concern has been with the interests and needs of retired investors who want to maximize the functionality of their investment accounts while minimizing their efforts to maintain those accounts. Disclaimers This article is for informational use only. It is not intended as a recommendation or inducement to purchase or sell any financial instrument issued by or pertaining to any company or fund mentioned or described herein. All data contained herein is accurate to the best of my ability to ascertain, and is drawn from the Company’s Prospectus, Statement of Additional Information, and fact sheets. All tables, charts and graphs are produced by me using data acquired from pertinent documents; historical price data from Yahoo! Finance . Data from any other sources (if used) is cited as such. All opinions contained herein are mine unless otherwise indicated. The opinions of others that may be included are identified as such and do not necessarily reflect my own views. Before investing, readers are reminded that they are responsible for performing their own due diligence; they are also reminded that it is possible to lose part or all of their invested money. Please invest carefully. 1 ” Major Changes In My Retirement Portfolio .” 2 My thanks to the reader, Edwardjk , for mentioning the portfolio strategy – very interesting, and worth a long-term look. 3 The main website to see is Browne’s own site, harrybrowne.org , although Mr. Browne died in 2006; a record of the Permanent Portfolio’s performance is here , but it only covers from 1970 – 2003. A concise description of the strategy can be found in the article ” Harry Browne s Permanent Portfolio: When You Can’t Afford to Lose Money ,” by Jason Jenkins. A more general discussion is presented by J.D. Roth in ” Fail-Safe Investing? Harry Browne s Permanent Portfolio .” Bogleheads.org , a site catering to Jack Bogle enthusiasts, has blogs such as this , as well as discussion groups. 4 Roth, here . 5 For his part, Browne used mutual funds in the original formulation of the portfolio, and only added one mutual fund to his list of suggestions in the 2001 update. 6 Fail-Safe Investing: Lifelong Financial Security in 30 Minutes, Harry Browne (St. Martin’s Griffin, 2001), p. 160. 7 BIL, which seems particularly useful in the capacity of “money-market” fund, was not issued until May 23, 2007; to bring that part of the portfolio back to January 1, 2005, I used SHY. Since neither SHY nor BIL actually grows in value much, there is little difference between the two. Browne’s intent in specifying cash, however, was to limit exposure to interest rates, and BIL is better suited to that purpose. 8 This keeps the weight on the S&P 500, and U.S. stocks in general, while adding a few foreign equities through DGT; this strikes me as keeping in the spirit of diversified investing while staying faithful to Browne. 9 I used RSP for S&P 500 coverage because – from previous study – I knew that it would outperform SPY. See my ” Guggenheim s RSP: Equal Weight Or Dead Weight? ” 10 The shift from BIL to SHV (which was made because SHV does pay an occasional distribution, while BIL has only paid one in the past four-plus years) shouldn’t matter. Neither fund undergoes any noticeable change in value in the course of the trials. Distributions paid over the past five years have been barely noteworthy. 11 The timing is the main reason why I see GLTR as being the issue, but it is also likely that DGT and/or QQEW played a role. 12 And only HBPP3 was rebalanced in 2012; HBPP2 would not get rebalanced again until 1 July 2013 – the last rebalancing that was needed for any portfolio. 13 ” Barrick Stays Bullish as Gold Prices Head for Third Annual Drop ,” Bloomberg. 14 ” Gold, silver set for more pain into 2016: poll ,” Reuters. 15 ETF.com . 16 ETF/R-A, for those who are interested, was discussed in my article; it contains the changes in ETFs that were made, but did not include the capital injection from accrued dividends that were part of the actual portfolio. 17 The rather severe drop shown just before 2015 is the result of a nearly $800.00 loss (-7.92%) on 10 December that was offset by a dividend payment of $3.17/share ($733.80 total, for this test). The fund’s current yield of 8.21% may be somewhat misleading – historically, its yield has been much less than $1.00 per share per year. 18 PERM’s yield is currently 0.98%. 19 I am not a tax expert. Each individual should consult their own tax advisor concerning the taxability of their investment income. 20 Investment companies differ in their policies about transaction costs and fees for services. Consult with your banker and/or broker to determine how your activities are charged.