In your circumstances a larger emergency fund is warranted- but did you stop investing to build it up, or did you continue investing AND build up that emergency fund?

Dylan,

I’m not against emergency funds- I think they are quite useful. However, I don’t think they are so immediately necessary to defer investing for any significant period of time. I even had a $1000 emergency fund before I started investing. That is only because I didn’t realize at the time how important it was to start investing as early as possible.

-Rick Francis

The 6 month number usually refers to 6 month’s worth of essential expenses, not 6 month’s worth of income. But like Mike said, it doesn’t have to be 6 months. Even a 3 month emergency fund will cover most emergencies. It does not have to and shouldn’t take years to build a reasonable cash reserve for emergencies.

Whatever method you use to determine the appropriate size for your own emergency fund, I still think establishing an appropriate emergency fund should take priority over long term savings for two primary reasons.

First, emergency funds are to meet short term needs. It generally makes sense to fund short-term goals before long term ones (I say “generally” because you do have to consider things like employer matches). Since the timing and size of emergencies are unknown by definition, you may not have a lot of time to save up for an emergency.

Second, if building an emergency fund sounds like a sacrifice now, how much more of a sacrifice do you think coming up with that money will be in a real emergency?

Sorry if I wasn’t clear enough, the 6-month number is just my number. I absolutely agree that for many investors, the number should be lower (perhaps far, far lower).

A big part of the reason mine is so high is that I’m self-employed, so my income can fluctuate rather dramatically from month to month.

Thanks for helping to highlight that particular aspect of the topic.

I have to disagree about building up a 6 month emergency fund before starting to invest especially for young single investors- it will take a lot of time it would take to build up that fund and compounding makes that a very costly wait.

Say a person saves 10% of their salary then they need 0.5 year * 90% =45% of their yearly income for their emergency fund. That will take 45%/10%/year => 4.5 years!

Say a 22 year old working their first job decides to build an emergency fund before investing so that they have 38.5 years of investing instead of 43 years of investing before retirement. Assuming an 8% CAGR and fixed yearly amount invested the cost of waiting is 31% of their portfolio!

That is a huge cost- and if they miss out on an employer match it is even larger. I agree that having an emergency fund is very useful- but not investing for years is too costly. I think it makes more sense to work on both simultaneously. Alternately save like mad for a few months and then invest.

-Rick Francis

The idea is to get the higher interest rate of a CD, but have at least one month’s living expenses readily available when you need it in true emergency without paying any penalties.

I agree that one shouldn’t go nuts with optimization. But I think being as informed as one can reasonably be is a good thing. The difference between a 6 month EF and a one year EF can be large and have major consequences for one’s future earnings. I think it is possible a seat-of-the pants estimate can be off by that much.

It is good to be conservative. One can’t really know how conservative one is being without analysis. When it comes to EF (as opposed to AA), there doesn’t seem to be that much analysis out there, just ballpark guesses and “do what helps you sleep at night” advice. That’s fine. But I think one can do better.

One can still use the “sleep test” even knowing the analytical results.

That’s all I’ll say about this. I’m not a crusader about this issue. If I cared enough about it I’d cook up an estimation technique myself. It is just a point about the “literature”. I think it is missing something on EFs. (And if I am wrong, I hope someome will alert me to a source.)

Bonds are meant to provide a safety net and balance the risk of equities. An emergency fund does the same thing for our income and unexpected expenses. While it may not always be the mathematically optimal choice, it provides a peace of mind that’s worth much more than that extra return in your portfolio.

Yeah, that’s the key assumption. But one could do a study that illustrates the trade-off for various assumptions. Maybe if nobody else does it someday I will.

It has always struck me as odd, though, that otherwise empirics-based and reasoned investors will just assume that an N-month EF is correct. It may be conservative (which is good), but it may be so by a large factor. Or it may be inadequate. Who knows?