Why an emergency fund is a bad idea in one single chart

Ever since we posted our view on emergency funds, we have been thinking about a succinct, straightforward but also scientific way to debunk that bad, bad, bad advice that investors should hold large amounts of cash in a money market account. Here’s one try:

There is nothing wrong with reducing risk as a response to facing a future cash flow uncertainty. We personally don’t feel that urge right now because we currently have enough income flowing in to counter such unpleasant surprises. But not everybody has that luxury. People with high risk aversion, whether it’s because of cash flow risk or just general uneasiness about equity volatility, should certainly reduce their portfolio risk level. But it’s very likely they shouldn’t do so by holding large amounts of cash in a money market account.

It all comes down to the efficient frontier analysis. This frontier is efficient in the sense that for any point on this line there exists no lower risk portfolio with the same expected return, see the dark blue line in the chart below:

EffFront with Emergency Fund
Efficient Frontier and why the Emergency Fund shouldn’t be in a money market account

If you find yourself on the efficient frontier past the tangency point (see above), one can easily show that reducing risk involves no cash holdings, but rather keeping all of your portfolio in risky assets. Specifically, you simply move along the efficient frontier and into other risky assets with lower risk and more diversification, e.g. bonds.

Mixing in cash (earning the money market interest) would mean we move along the red connecting line between the money market return point (zero risk, very low return) and the initial portfolio. But by definition, this line cuts into the interior of the efficient frontier. You reduce expected return by more than you have to. If instead you walk along the efficient frontier you can reach the same expected risk level but with a higher expected return (or alternatively the same expected return but lower risk) than the portfolio with emergency cash. Of course, you still give up some expected return, that’s the opportunity cost of lower risk, but at least you avoid the efficiency loss of the money market fund.


EffFront with Emergency Fund calc
Numerical Example

In the numerical example, see table above, our investor is no longer comfortable with a 100% equity portfolio. Instead of keeping 20% in cash, thereby reducing expected risk to 12%, the investor could move into 10y government bonds with a higher return than cash and even a little bit of negative correlation with equities. You can reduce the equity share by just under 20% and still maintain the 12% risk target. The expected return is about 0.30% higher than under the cash scenario. 0.30% is not trivial when we consider that this is simply from reshuffling 20% of the portfolio!


For the defender of the emergency fund, if the argument has to do with risk mitigation we agree that an (almost) all equity portfolio like ours might too risky. But holding large amounts of cash is not the right solution either. Investors should move along the efficient frontier, not to the interior of the efficient frontier!

Some Caveats:

  • This doesn’t apply when you’re not past the tangency portfolio. But few people are. With enough diversifying assets the tangency portfolio has a very low expected return and low risk and we can almost be guaranteed that most investors are beyond the tangency point. In the numerical example above the tangency point has about 60% bonds, 40% stocks, generating 3.7% expected return, much less than most people target. With more assets in the mix this could be even lower (shifting the frontier to the left and the tangency point in a southwestern direction).
  • All of this analysis doesn’t apply if you are already willing and able to take the tangency portfolio and scale it up through leverage, thereby effectively shorting cash. See our post here on why you want to do that. In that case, you certainly don’t want to hold cash in an emergency fund. You would merely short less cash if you want to reduce risk. The same result holds but the analysis is sightly different.



30 thoughts on “Why an emergency fund is a bad idea in one single chart

  1. I agree with you that many people put their money in money market accounts or CD’s, much more than they need beyond an emergency. For example, in Canada 40% of Canadians put 70% of their retirement savings in GIC (or CD’s in the states) and money market accounts in fear of the stock market going down and losing their money . This 70% refers to retirement savings, so we are talking about monies that will not be touched for a minimum of 10 years down the road. That is pretty scary. Keeping a minimum of 3 months of life expenses in a money market account or GIC in the event of an emergency is prudent because if the market goes down right when you need the money and all of your funds are in risky equity investments, then you are hooped. However, I think many people keep a lot of money in ‘safe investments’ like money market accounts out of fear of loss and lack of investing knowledge, not because they want to. Like anything else, the more you know about investing, the less scarier it is. Good post.

    1. Thanks for stopping by again and thanks for the international perspective from North of border.
      The fear of stocks dropping right around the time when you need the money is real. But in the last few episodes of sharp stock market drops, bonds went up (US government bonds are a safe haven asset and appreciate in crisis periods) so the only thing better than 3 months worth of expenses in a money market fund is having 3+x months worth of expenses in the bond portfolio due to higher bond yields and negative correlation between bonds and stocks. 🙂

  2. Very interesting point. Is there a reason why the stock weight along the Efficient Frontier stays above 80% when you target 12% risk? Thanks!

    1. Thanks for the question! With 80% stocks and 20% bonds you’d actually have less than 12% risk. That’s because stocks and bonds are negatively correlated. So, you can reduce the stock exposure by a little bit less than 20% and keep more in expected return. 🙂

  3. I’d note that you can combine emergency fund with fixed income allocation, and not suffer. Using your “emergency fund” to chase bank sign up bonuses can give a competitive or even superior effective yield. I’ll be getting over 2% with bank savings accounts this year counting bonuses, seems like best of all worlds (got $500 bonus in 90 days on 50K from Capitol One for example, a deal that is now expired). I’m also baffled at the return on cash being 0.375%, even without bonuses it is easy to get 1% in an FDIC insured high-yield savings account at a number of places (Synchrony is 1.05% currently).

    1. As cash yield I was using the Fed Funds Target Rate. True, actual money market rates can be a bit higher and that higher cash yield would roll up that tangency point, but not by much.
      I could have also used a 30Y Government Bond with significantly higher yield (2.3% vs. 1.5%), but for the sake of the example I picked the 10Y.
      With a 1.05% cash yield and still the 10Y bonds, the Emergency fund would lag behind by 0.17% p.a. (5.98% for efficient frontier, 5.81% for Emergency Fund)
      With 30Y bonds and higher yield the Emergency Fund would have lagged behind by 0.29%. In a $1,000,000+ portfolio that’s a lot of money. True, you can “chase” sign up bonuses here and there but 0.29% of a million dollar portfolio is $2,900 p.a. Not sure how to compensate for that.

  4. if you are self-employed and have ‘lumpy’ income, I’m a pretty big fan of putting a big chunk of your emergency fund into a short-term corporate bond index. 2% yield and only dropped 5-10% during the Financial Crisis. Glad to see someone agreeing with my “cash is bad” stance.


    1. Welcome to the blog, Casey! Agree, bond funds are much better than CDs or money market funds. Also, during normal fluctuations (outside of Global Financial Crisis) there is very little correlation with equities.

  5. Very cool approach to the subject. It will resonate with those that are able to weather an actual emergency, but considering most Americans don’t even have a couple of grand put aside for when shit hits the fan it seems very theoretical in most cases.

    1. Thanks! Our approach might have followers among those who already have many months worth of expenditures saved and who may not need additional cash sitting around in a MM account.
      But I also think that for people who just start out as savers, maxing out your savings in productive assets first isn’t such a dumb idea. Why save in a low interest account when you can use the power of compounding in your productive investments starting at a young age? Again, if people don’t like risk, mix in some bonds (but no more than 10% if under 35), but the inefficiency of cash still holds. Whether your portfolio is $10,000 or $10,000,000.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.