I recently told some friends and co-workers that I am planning to write a personal finance blog, and with that began a deluge of questions. What should I invest in? What is the best retirement savings vehicle? Is it better to buy or lease a car? It was actually kind of amazing to see these subjects which are usually off-limits being discussed more freely. That made me really happy, and is certainly one of my goals in writing this blog.
These questions also gave me lots of ideas for future topics to cover here. Today, I’d like to talk about modeling finances for retirement, a subject that came up with one of my co-workers. He asked me to take a look at the spreadsheets he had created to model his finances from now through retirement. First, I was super impressed that he had spreadsheets at all. In my experience, anyone who has financial spreadsheets is probably in good shape, since this indicates a general awareness of financial issues. Of course I agreed to look at his spreadsheets, with the caveat that I am NOT in any way certified to do this and am basing my analysis off of what I learned in my singular personal finance class and my personal experience.
Do you know what my first reaction was upon seeing my co-worker’s spreadsheets? Jealousy. Those were some kick-butt spreadsheets. They were pretty, they were organized, and they were thorough. They accounted for the myriad factors that traditionally go into a retirement savings calculation: savings rate, current savings, projected future savings, inflation rate, rate of return on investments, withdrawal rate in retirement, etc.
As we dug into the numbers, we talked about all of the assumptions that went into the calculations. Should the rate of return on investments be 3%? 5%? 7%? Between 1928 and 2014, the S&P 500 returned, on average, 11.53% per year. 3%, 5% or 7% were all conservative estimates of return based on past performance. But what if your timing happens to be awful and you need to withdraw money during a downturn? Or what if the S&P 500 crashes? You can see how the “what ifs” could go on forever.
The same uncertainty applies to all of the other assumptions that go into a model of this kind. Inflation could be crazy high or could remain at the recent historical lows. You might get a huge raise or win the lottery. You might lose your job and have to switch to a lower paying industry. You might have kids. You might not. You might buy a house, and that house might go underwater, or it might appreciate substantially. Some of these things we have the ability to influence (I wouldn’t say that we really control anything) and others are completely outside of our impact.
How do I model uncertainty?
All of this uncertainty makes a financial model, well, worthless. You could drive yourself crazy modeling every possible outcome, but I think that is just going to make you depressed. Plus, what are you going to do if you know that if the markets crash and you get fired and your house is underwater you won’t have enough money to retire? Realistically, it may not even be possible to save enough to meet your theoretical savings goals. I’m sure you could build a model that would tell you that even if you saved 100% of your income you still wouldn’t have enough to retire at 65. What do you do with that? Do you live like a miser and make yourself miserable today for some future potential benefit in a scenario that may or may not happen? Do you quit a job you love to become an investment banker? I say no.
I would like to add a caveat here, and that is that financial models for retirement are more useless the further you are away from retirement. My co-worker is 33. His isn’t married and doesn’t have kids or a house. But he wants all of those things. He is at the beginning of his career, and it is hard to predict how much money he will make every year into the future, how his expenses will change with the addition of a house or a family, or how the world will change in his next 30+ years in the working world.
Financial models for retirement make a lot more sense when you are actually approaching retirement. For example, my dad recently retired and my mom is not far behind him. Before my dad retired, he did all kinds of calculations to model how long their money would last under different scenarios, where to put their money so that it was likely to grow but still remain insulated from too much loss, how much of their savings they could reasonably spend down each year, etc. These things make sense because my dad knew what he was working with and the models helped inform the exact timing of his retirement. The variables were fewer therefore the models were more useful.
The irony here is that if you wait until you are able to model the variables with a reasonable degree of certainty to begin saving for retirement, you’re too late. It is good that my co-worker is thinking about retirement now, even though it is really hard to determine how much he really should be saving for it.
If my model is crap then how do I know how much to save?
Given my lack of faith in financial models, I told my co-worker the same thing I told myself several years ago. I was 25 and taking a personal finance class as part of my MBA. We had to determine how much money we would need to achieve certain goals (buying a house, paying for kids’ education, retirement, etc.) and then how much we needed to save in order to reach those goals. I worked through all the models and was left with a feeling of despair. On my current salary, there was no possible way I could meet all my goals, even with a rosy economic outlook.
But then I realized that there were so many things I was leaving out. My salary is going to increase (theoretically) ever year. At some point, my savings goals will hopefully be shared between two salaries when I get married. My condo might appreciate in value, leading to a windfall when I sell. Heck, college might be free by time my kids get there (fingers crossed!) These are the potential upsides that would lighten the savings burden. Of course there are downsides: I could have a medical emergency and have to take on debt to pay it off, I could get fired and fail to find a new job, college costs could go through the roof, etc.
I realized that none of this was helping me, because it just wasn’t giving me useful information. If a model tells me that I need to save $3,000 a month to pay for my hypothetical future child’s college education, but my take home pay is only $2,800 a month, what do I do? I decided that all I could do was my best, and that the models weren’t helping, they were hurting.
Save as much as you can
So I abandoned my models (after turning them in – I didn’t care about them but the teacher sure did!) in favor of this simple idea: save as much as you can starting as early as you can. Make prudent financial decisions, avoid lifestyle inflation, and acknowledge that the future is unknowable and no amount of planning or modeling is going to change that.
This is the strategy I employ myself and also the advice I gave my colleague, because it was clear that he was already aware of the importance of saving (as evidenced by his healthy net worth and extensive modeling efforts). This strategy probably wouldn’t work well for someone who isn’t convinced of the merits of saving and is more interested in instant gratification. For someone who likes to spend their whole salary (or – the horror! – more), putting some numbers into an online retirement calculator might be a wise awakening. For that person, setting a goal of putting aside $X per month is probably useful.
For us savings nuts, goals like that are more anxiety producing than useful. I am pretty good with delayed gratification. I worry about my future self quite a bit and would rather suffer now than suffer later. It is hard enough for me to balance enjoying my life today and feeling secure in the future without a model giving me a doomsday scenario. So I’m doing my best, and trying to be happy with that.
Do you model retirement savings? What tools do you use to do so?