thumbs down

[Hey guys! Have a feisty one for you today from fellow blogger ERN at Some of these you may disagree with, as do I (no emergency fund? Blasphemy! ;)) BUT I always love hearing other peoples’ perspectives, and Ern here can surely put up a convincing fight. Regardless of where you stand though, I hope you’ll find some takeaways here to help you in your own journey.]


As members of the FIRE (Financial Independence, Retire Early) community, we are all nonconformist. We scoff at conspicuous consumption of our neighbors in their full-size SUVs and McMansions, and instead budget with a very sharp pencil (spreadsheets are fine, too) and grow our stealth wealth, one dollar at a time.

But the FIRE community is far from monolithic. For example, I consider myself nonconformist even inside our nonconformist movement! I hold some pretty strong non-consensus views on many financial topics. I’m a rebel among the rebels, imagine that!

Here’s a list of my seven favorite disagreements with the personal finance consensus:

#1: The 4% Safe Withdrawal Rate is safe! (No it is not)

The Trinity Study and the resulting 4% recommended safe withdrawal rate have become the gold standard when designing a savings target (25x annual spending) and the eventual withdrawal strategy. Even J$ here uses it. There are (at least) two problems when extrapolating the Trinity calculations though to the early retirement community.

First, instead of a 30-year horizon in the Trinity Study, the average early retiree faces a 50-60 year horizon. That wouldn’t be a problem if the success criterion in the Trinity had been capital preservation, i.e., preserving the purchasing power of the initial portfolio value for 30 years. If your capital lasts for 30 years, then why shouldn’t it last another 20-30 years? But for the Trinity Study, even a portfolio value $0.01 after 30 years is considered a success. That one cent will not last another 20-30 years!

To gauge how much of a “haircut” we should apply to the 30-year safe withdrawal rates when extending the retirement horizon, let’s do the following thought-experiment.

Imagine we have a $1m portfolio that generates 4% p.a. of real return in the long-term, but we got unlucky and suffer 10% losses for the first three years, followed by three years of 10% gains and then 4% returns after that (assume all returns are real and all withdrawals adjusted for inflation). That’s not even a worst-case scenario, more of a garden-variety bear market.

Our portfolio can then sustain a withdrawal rate of about 4.27% p.a. for a 30-year horizon. But over longer horizons that number will be significantly lower: Only 3.26% over 60 years, a whole percentage point less and that’s all because of the longer horizon!

safe withdrawal routes time horizons

[Figure 1: Path of portfolio values over 30, 40, 50, and 60-year horizon with different withdrawal rates, targeting capital depletion.]

The second objection is that today’s expected asset returns are significantly lower than the average returns observed since 1926. The Trinity Study averages over many decades with different financial market regimes. Sometimes the stock market was overvalued, sometimes undervalued. Sometimes bond yields were very high, sometimes they were very low.

Does anybody else see a problem with that procedure? Relying on safe withdrawal rates that are averaged over the past 90 years is a little bit like calculating the probability of getting into a traffic jam by averaging over the entire 24 hours of the day. That may not be the most informative figure if you already know that you will be driving during rush hour!

For today’s retirees, only today’s market conditions matter, not the averages over the last 90 years. And more than seven years in the current economic expansion, equities look more expensive than over the last 90 years (measured by the CAPE ratio) as I show in this blog post. Bonds don’t look too hot either. Since 1926, the real, inflation-adjusted bond returns were solidly above 2% for government bonds, and even above 3% for corporate bonds, but today’s yields are far lower. Lower bond yields and expensive equity valuations support only significantly lower safe withdrawal rates than 4%.

If 4% doesn’t work, then what’s the alternative? Personally, I plan to start with a lower withdrawal rate of 3-3.25% out of our equity portfolio to account for today’s expensive valuations and the long retirement horizon. Having income-producing assets with a less than perfect equity correlation is also a good idea, so, I started moving some of my investments into rental real estate. Owning rental properties with a 4%+ rental yield (after all maintenance and repair costs!) I can probably push the overall withdrawal rate to 3.5%.

This number, though, is not set in stone. If the CAPE ratio were to drop to a more normal level, say, under 20 (currently around 27) I could move the withdrawal rate closer to 4% again.

#2: Robo-Advisers are great! (If by “great” you mean “expensive”)

You can’t read through the personal finance blog world without finding glowing endorsements of Robo-Advisers, like Betterment and Wealthfront. I beg to differ though, and recommend Robo-Advisers only to my most financially disinterested friends.

But folks in the FIRE community? We are the masters of budgeting and frugality, financial hackers (not hacks) and always on our toes to find new ways to eke out a basis point (0.01%) of return. Why would we want to throw away anywhere between 0.15% and 0.35% in fees p.a.?

Robo-Advisers do nothing magical and are nothing but expensive gimmicks. You can go to their website, find out their recommended asset allocation and simply implement it with inexpensive index funds at Vanguard or Fidelity. But do you even want to use their recommended asset allocation? You could do significantly better by further hacking their recommended allocation, for example moving ETFs with a high dividend yield and taxable bonds from taxable to tax-deferred accounts.

But what about tax loss harvesting? Tax loss harvesting is a neat tool that can lower your taxable income by up to $3,000 p.a., so folks in high tax brackets can save $1,000 or more on their annual tax bill. In taxable accounts, simply sell your underwater investments, i.e., equities, ETFs and mutual fund shares that have a cost basis greater than their current value, and the loss can be used to offset up to $3,000 of ordinary income per year. Check Bogleheads for more details.

Robo-Advisers can do this process really well and in a systematic and automated fashion. But apart from the fact that we find Tax Loss Harvesting overrated (see #7 below), it’s also something that most investors should be able to do themselves. Here’s my guide to becoming your own homebrew Robo-Adviser. Do it yourself and save thousands of dollars over the years!

Finally, under no circumstances should anybody ever shift an existing brokerage account with sizable capital gains to the Robo-Advisers, because they might first liquidate your holdings to purchase their recommended ETFs.

And that’s true whether you are an experienced financial hacker or complete finance novice. I’ve saved $42,000 by not switching to Betterment! The tax bill could be so excessive you’ll never recover the loss even under the most optimistic assumptions for harvesting future tax losses.

#3: Everyone Needs an Emergency Fund! (Not everyone)

The almost universally accepted wisdom is that not only do we all need an emergency fund, but building that emergency fund is the number one priority of personal finance, taking higher priority than saving for retirement. Some even want you to start an emergency fund before paying off high-interest credit card debt. That’s nuts!

I laid out our plan for having no emergency fund (featured on RockstarFinance on 5/25/2016!) and two follow-up posts as well to debunk some of the common arguments in favor of the emergency fund. With our savings rate of 60%+ we are usually able to finance all expected and unexpected expenses out of our current cash flow.

Just to be sure, I’m not saying that anyone should forego savings altogether. An emergency fund is better than having no savings at all. I merely take offense in keeping large amounts of money in unproductive, low-interest money market accounts. I consider our entire portfolio our emergency fund and like to put our hard-earned dollars where they can work harder for us: in equity and real estate investments.

Am I not afraid of having to dig into an equity portfolio right when the market is down? Yes, to a degree, but as a passive investor, I try to stay away from timing the market. Keeping cash on the sidelines for fear of a market decline exactly coinciding with a cash flow need is market timing on steroids!

#4: Use bonds to diversify equity risk. (Don’t get your hopes up too high!)

What was the correlation between an all-equity portfolio and a portfolio with 80% equities and 20% bonds over the last ten years? 0.998. For all practical purposes, that’s a correlation of one.

To be sure, the 80/20 portfolio has a lower volatility; pretty much exactly 20% lower. But that volatility reduction came from the lower equity weight and had little to do with bonds reducing risk. In fact, simply keeping the 20% in a money market account would have achieved that exact same volatility reduction. That said, bonds delivered very nice average returns over the last few years, so the appeal of an 80/20 stock/bond portfolio came mostly from better returns than investing 20% in a money market account.

But that could change in the future; yields have been heading higher since November 8th, and bonds could have a rocky road ahead of them. In addition to having little diversification, they could also lose their potential to boost returns.

#5: Bonds are safer than stocks. (Depends on the horizon!)

Of course, stocks tend to have higher daily, weekly, monthly, even annual volatility than most bonds. No discussion about that. So, for folks with a very short investment horizon stocks may seem unattractive. But over a 50-year investment horizon, we should weigh the short-term volatility with the long-term sustainability of funding our expenses in retirement.

There have been extended periods of very poor bond returns. For example, there was one 80-year window (!) of zero real (inflation adjusted) bond returns from about 1900 to 1982. True, equities can be volatile in the short-term. You see overreactions both on the upside (late 1990s) and downside (2009). But equities normally return to a long-term trend growth path within a few years, see chart below.

Bonds, on the other hand, can move sideways for many decades. That’s poison for the retiree who relies on a nest egg to last for half a century! And what’s worse, the spectacular run of good bond returns that started in 1982, might be overdue for a reversal as has happened with the other two bond bull markets in the late 1800s and 1920-1940. I’m not saying that this will start now, but bonds did get crushed after the spike in interest rates since November 8th!

bonds riskier than stocks
[Figure 2: Which one is the risky asset now? Cumulative real total returns (CPI-Adjusted) of the S&P 500 and 10-year U.S. government bonds. January 1871- December 2015.]

So, to stay with our “financial rebel” theme, just like our American Founding Fathers preferred “dangerous liberty over peaceful servitude,” this writer likes dangerous equity volatility more than running out of money for sure with low volatility bonds!

#6: Cash serves as great bear market insurance in retirement. (Think again)

The idea sounds almost too good to be true: simply keep a few years’ worth of expenses in a money market account, enough to fund expenses during the occasional bear market, and we never have to worry about market volatility. Unfortunately, it is too good to be true.

The first issue is opportunity cost. The cash cushion is a little bit like leaving the house every morning wearing a Robin Hood costume, just in case I might come across a fancy-dress party that day. That strategy works beautifully every year in late October, but it would be a bit of a burden for the rest of the year. Likewise, the opportunity cost of carrying too much unneeded cash when there isn’t a bear market can compromise the average portfolio return. Remember, timing Halloween parties in late October is easier than timing the next bear market!

But even if you do get lucky and have the cash cushion ready to go at the onset of a bear market, it’s not going to make much of a difference. If you were unlucky and started your retirement too close to the market peak in 2000, then with or without the cash cushion your retirement portfolio would have been seriously compromised; we are talking about “going back to work” compromised! True, the portfolio with the cash cushion would have been slightly less compromised than an all-equity portfolio, but still seriously underwater. The cash cushion is no panacea because you will have to replenish the cash cushion right around the time when the stock market rallies again.

One way or another, opportunity costs will catch up with you! It’s very simple: there is no true bear market insurance. A good start would be using a lower withdrawal rate (see #1 above) and being prepared for plan B, plan C, and all the way to plan J (Money).

#7: Tax loss harvesting saves you a ton! (It’s nice, but you won’t get rich)

As mentioned above, I do the tax loss harvesting by hand without any help from the Robo-Advisers and my guide on how to be your own Robo-Adviser has a lot of the details on how to do this right.

Did I become rich from tax loss harvesting? Absolutely not! The claims of 0.7% or even north of 1% of additional annual returns from tax-loss harvesting claimed by the Robo-Advisers are vastly exaggerated. If you are like me your taxable portfolio is the result of many years of regular investments. So, even a bear market with a 20%+ drop right now would result in much less than 20% harvestable losses. I have some tax lots that would require a 50%, even all the way up to 70% drop in the stock market before generating even a single tax loss dollar.

My personal estimate is that I have generated less than 0.10% in additional annual return from tax-loss harvesting so far. Going forward, especially once I retire, the benefit will likely go away altogether.

I put together some calculations on how much extra return you’ll likely generate depending on different marginal tax rate scenarios and portfolio sizes. Under most realistic assumptions, the benefit is even lower than the 0.15% to 0.35% Robo-Advisers fees. It’s a benefit large enough to not leave on the table, but don’t get your hopes up too high. And don’t pay anyone 0.35% p.a. to do it for you, either!

I’ll now be taking all questions and concerns below 🙂

About the author: Mr. ERN runs the EarlyRetirementNow blog where he writes about his path to early retirement, planned for early 2018. He currently works for a large asset management firm and holds both a Ph.D. in economics and the CFA (Chartered Financial Analyst) designation.

[Top photo cred: hobvias sudoneighm]