• 08 Nov, 2025 *

I had never budgeted before. I have always been a naturally frugal person, wracked with shame about spending amounts that are reasonably within my income bracket, and so the whole “set up a spreadsheet for your savings!” thing just sort of came to me. But as my income grew, my deep mindset about saving money did not, and I came to the realization that I should budget not from the “Wow, I need to get my life together” direction and more from the “I’d like to not feel weird about getting an appetizer” direction.

The problem was that most mainline budgeting systems are extraordinarily convoluted and didn’t ultimately align with what it feels like as a human to spend money. Some of them, like the envelope method, are based around buying most things in cash which, lol. Others are extraordinarily bulky spreadsheet demons that required me to sit down with my bank statements and figure out “how much am I willing to spend on lunches for a month? What about entertainment? What even counts as entertainment?”

I found two problems with every system based on the way that I (and I assume a lot of people) actually think and feel about money.

The systems are based around months, which is how rent and taxes work, I guess, but is not really applicable to how one taps their card for Taco Bell. A month is an inconceivable amount of time for someone to actually think about. Can you imagine if other habits were recorded on the order of a month? Fitness plans the world over telling you that you should do 300 reps of bench press per month. Dietitians telling you to get 850 grams of fiber per month. Something something hyperbolic temporal discounting if you want to get nerdy about it.

And the systems are based around rigid buckets, which is not how the money in my bank is separated. Let’s say I don’t go out very often for drinks, but I’ve got a friend who occasionally gets a big jolt of energy to go out and invites me. I don’t know when this is going to happen. I already estimated my entertainment/fun bucket for this month, and this is going to throw it off. I suppose I can just not have drinks for the rest of the month, even though I would rather cheap out on the next few lunches to make up for it. Being rigid to the bucket is no way for someone who lives in a world full of surprises (hint: all of us). Once again, something something stated vs. revealed preferences if you want to get nerdy about it.

I eventually came up with a basic way of figuring out what my daily discretionary spending is. It has evolved far past the basic idea as I’ve used it, and it has revealed some added interesting mechanisms and benefits that I had not originally expected. I call it Float.

The Basic Concept of Float

The basic concept of Float is simple. If you, like most people, get paid biweekly, take 2 of your paychecks net of taxes. You subtract out all the things that need to or are going to be paid whether you like it or not: Rent, bills, debts, subscriptions, retirement contributions, and (importantly) actual savings. You divide this number by 28 (the amount of days in a biweekly period). This is your Float.

That’s the system at its cleanest. You get one number that you can spend, no shame at all, per day. There’s no weird time-discounting thing going on about “what if something comes up next week?” You do discretionary paying for things like, what, 4 times per day on average?

You can stop there. It works pretty well as just a straight guideline. Or you can, like me, take it way farther to see how well you can complicate a system if it starts out simple. Starting with a basic level, two very basic questions jump out.

“What happens if I don’t spend the whole float?”

I just give it to the next day. It “floats” (ha) on over—whatever was unspent. This is nice because it typically renders down the natural ebb and flow of weekdays vs. weekends of going out and doing fun things. I spend less than my float M-F, and when Saturday rolls around, I’ve got a huge number I can spend that day, no strings attached. There are some extra mechanisms that can be applied to this rollover system that we’ll touch on later.

“What happens if I have to/accidentally spend above the daily float?”

In the case of actual necessary payments, I just use my savings. A new tire is not the same thing as a sandwich. But if you got one too many margaritas and forgot to check your float, you can just take it out of tomorrow’s float, or the next 2 days, or whatever. Divide up the excess by however many days you want to “take debt” off of, and your number is just lower for those days.

This is the part of the system that has the biggest threat for a certain class of people. You can just keep spending over your float and taking it out of the next week until you end up in a cycle where it is impossible to live on the amount you technically should be spending. There’s not a super nice way to deal with this. I just suspect that there is some portion of the population who can’t really use any system that has an internal debt component. But if you’re reading the personal finance system of a niche blog, you probably are not that sort of person.

This sort of “float debt” is an inherent soft-commitment device. I have plenty of times gone, “I want X thing that is over today’s float,” and when I purchase it, I actively acknowledge that there’s not some dollar amount that is going to go up on a screen and become a problem for me a month from now (a fake amount of time for humans, as we established). Before making the purchase, I know that, hey, I’m gonna be a bit stingier for the next 2 days. And if I’m really stingy tomorrow, the day afterward is entirely normal due to rollover. It all comes out in the wash.

The personal way that I implement this is with a little pocket notebook. You can use your phone’s notes app. Just round up to the nearest dollar, or $5 if you’re working with bigger amounts so you don’t get weird and angsty about cents on receipts, and just subtract and add. An extraordinary feature of this system that originally happened just due to me being lazy is friction-based savings. Doing this basically means that I’m adding a 5-10% “sales tax” to myself just from rounding up. My account is larger than I expected because I’m actually spending less than I was accounting for. This is a bug-turned-feature.

Now you might do your calculation and realize it’s a bit tighter than you like. I did this originally too. It was still good to know the exact amount. There is another bug turned feature in this system I’ve been hiding, similar to the one about rounding, and you might have noticed it much earlier. This system automatically saves you 1 extra month of non-discretionary bills per year. We are building our budget around 28 day cycles, of which there are 13 per year. We are paying almost all of our bills on monthly cycles, which happen 12 times per year. As a side effect of the float system, we are putting an extra month per year of runway in savings. If your float is very very tight, you can math this out and remove this amount from the savings you are considering for the year anyways to find a better discretionary amount, but if you can get away with it this is a good scheme.

Why I Think This Works

I think what this fundamentally does is, from a sort of Meadows Systems Thinking perspective, to just treat money like a flow rate (which it is) rather than a set value, and give you information about it—in this case, “How much money can I spend, today?”

While you might not subscribe to the specific details, it seems to me that the theory of Hyperbolic Discounting has some legs. For those who aren’t aware, it is essentially the claim that human beings are pretty good at reasoning across time but not evenly: “Today vs. Tomorrow” has much stronger intuition behind it than “20 days from now vs. 21 days from now.” This is what I’ve been pointing to in my jabs that a ‘month is a fictional unit of time.’

Monthly budgets collapse entire weeks into “later,” which you likely will discount into irrelevance, or at least very far away from their actual value. I think there are two typical failure modes here.

The first is that you assume that today is significantly more valuable than the far future, leading to blowing all your money early in the cycle and then living extraordinarily scrappy at the end of the cycle. Insert the “day after paycheck: lobster dinner; day before paycheck: ice soup” meme.

The second failure mode here is the one I experience, in which you know that you have a tendency to discount the future, so you over-correct and never spend any money. This is the same thing as hoarding all your potions in an RPG because you assume the later bosses will be even more difficult, but then you never end up using them at all.

There is also some element of decision fatigue. Some budgeting mechanisms require you to preemptively categorize every possible transaction. I don’t know if I believe the whole “daily executive function budget” concept, but something adjacent to it feels right. Spending brainpower on categorizing things that are all coming out of the same flow rate feels useless, if not harmful.

The Float Debt specifically is an interesting method. You can overspend, in the same way you can put money on a credit card, but if we loop back around to the hyperbolic discounting thing, the fact that you have to take it out of the immediate future’s fund rather than some lump sum you’ve got ‘a month’ to figure out makes it a stronger selection and soft-commitment device. Future-you is a fictional figure who you expect to figure out the hard decisions to make up for what you want right now, whereas the late-night you and the morning you are much closer to the same being. (Thought experiment for readers: Can you imagine how much worse some people would be at drinking if they could bank their hangovers for “a month from now, with interest”?)

I think the really important part of Float is the substitution effect, which I didn’t even design for but is now the ‘killer feature.’

Because I’m now working with a single liquid budget rather than rigid categories, everything I do has to go through some sort of mental trade-off, which, surprisingly, the (at least my) brain is very good at, assuming clear framing.

Concrete example: It’s lunchtime, and I’m going to happy hour with friends tonight. This will be very fun, and I want to be able to not think about it too badly. Lunch doesn’t matter a bunch to me. When the option is an $8 burrito or a $15 bowl, I subconsciously consider whether the enjoyment of the bowl as compared to the burrito is greater than the enjoyment of an extra $7 drink tonight. It is not. I choose the burrito. I don’t have to really think about this, it just sort of happens now that I have the number in front of me. I don’t have to sit down and audit my values like I’m Marie Kondo.

You may have heard the financial advice about “It’s okay to spend money on what you value if you ruthlessly cut the things you don’t.” That feels quite correct, but it falls into the classic trap of stated vs. revealed preferences. If I make a budget ahead of time with that belief, I convince myself that I’m a guy who loves traveling and art museums and so I’ll cut out the nice meals (this is almost the exact opposite of my actual values). But over time with my Float as the focal point, I naturally adjust my spending downwards in the areas I don’t really care about to reach some equilibrium where a significant portion goes to things I care about on that specific day.

After a few months, you might realize you spend a lot of money on live music and not a lot on lunch even though you never really thought you were a “music person.” When the option presents itself, it is now not a “cost of ticket compared to my liquidity” or “cost of ticket compared to my entertainment budget which I misallocated,” it is “cost of a ticket compared to a day’s float.” It’s almost an internal prediction market about your own utility function. Float is your currency, spending decisions are bids, and the system converges on your real preferences.

Advanced Mechanisms on Top of Float

The Wipe

If you run a Float for a while and are a frugal person, you might find yourself with a stupidly high float left over. A few days of going a little bit cheap means you suddenly have, like, 3x your float as optionality on any given day. This can be good, but beyond a certain size, it becomes too large to meaningfully think through, almost like when an employer gives you “unlimited vacation days” so you take none of them.

You should probably set up a wipe method. You can format the wipe method in either days or amounts, i.e., “after 5 consecutive days in the green, I get rid of all of it” or “my daily float can never be higher than 2x my valued float,” and the rest gets hand-waved into the aether (i.e., your savings).

Downward Ratcheting

Related to the wipe, if you find yourself consistently spending lower than your float, you can just lower it. The hedonic treadmill of getting more money so spending more of it is pretty typical, but I have a theory that organizing it from the action backward rather than the other way around might make it easier. i.e., saving a larger percent of your paycheck might feel more of a loss because you’re removing the optionality of money you receive, whereas getting rid of “excess float” feels like you’re actively doing something charitable for your future self. Just a theory.

Dual-Float (For Real Sickos)

I’ve been toying with the idea of a dual-float architecture for a while now. A big problem is that some purchases that are financially reasonable still require multiple weeks’ worth of saving. Right now, if you have a wipe system on your excess float, you can’t get to that amount. Of course, you probably have the necessary liquidity to responsibly buy them just out of the extra savings from your wipe, but the fact is you don’t know how much you’ve put in, and the whole idea here is we want to make it clear what you can and can’t do.

So, we split Float into layers.

• X = Your actual observed burn rate. Track for a month or two. Just for this example, we’ll say, like, $50 a day.
• Y = Operating Float. Like 1.1 or 1.05 times your observed burn rate for flexibility. We’ll say $55 a day.
• Z = Your technical allowable Float. That singular number of the amount you can reasonably spend net of all the important stuff. We’ll say in this case it’s, like, $70 a day.

The gap between Z and Y can become a monthly discretionary capital budget, or just “Big Float,” if you will. That’s 28 * (Z-Y) per month.

In this case, 15 * 28 = $420/month of capital budget.

This is for durable goods, little treats—things that make your life easier in some sense. A big pool you can ball out on or save or just spread a bit here and there. New boots, plane tickets, whatever. It is important to denote this not as savings but as the Bank of Making Your Life Easier. It eliminates any shame related to absolute dollar amounts, even when it is proportionally fine. Spending $200 on shoes out of your savings, if you had an upbringing like mine, feels extraordinarily excessive and wasteful, but if you have that money, spending it is fine! You can also set a wipe for the monthly cap, maybe every 3 months to get yearly quarters. You could also take out of the monthly for your Float Debt, although I’ve never tried this mechanism.

A Concrete Example

You’re some late-20s, early-30s, tech-adjacent person living in a second-tier city. You make $85k a year. Post-taxes, we’ll say $63k or so. Biweekly paycheck of $2,430, multiplied by 2 for $4,860 per 28-day period.

Your fixed costs (that you pay from your take-home pay):

• Rent: $1,400 (1br)
• Utilities/Internet: $150
• Car Payment: $350 (you financed a civic or whatever)
• Car Insurance: $120
• Student Loans: $300
• Phone: $70
• Subscriptions: $50 (spotify, netflix, random shit adds up)
• Actual Savings: $400 (emergency fund, general long-term)

Total Fixed: $2,840

Your Float = ($4,860 - $2,840) / 28 = $72.14

Call it $72, or round down to $70 to be more conservative and have a nicer number.

On a default day, you might be doing:

• Coffee: $5
• Lunch: $15
• Groceries/Cook Dinner: $10
• Misc: $15

For a total of $45. This nets you $25 in positive rollover (using our $70/day conservative float). If you do that for 4 days (Mon-Thurs), your Friday float is now $170 ($70 base + $100 rollover). You can go out on Friday no problem.

If you find that you typical spend about 45 in an average day, you might consider lowering your actual daily float to 55 and then setting up a dual rate float. In this case, you’d get a (70-55)*28 = 420 dollar a month Big Float. Alternatively you could “steal” the savings you made by rounding down your initial calculation, leading to a Big Float of 476 dollar Big Float per month.

Ultimately, this system has been working very well for me for about 9 months now. It may not work for you, or may require some shift in the exact calculations to fit your mind better. I think if you ever alternate between being very thrifty to save and feel shame about how much you spent, float might work well for you just to give you a stronger sense of what the “speed limit of your money is”. If you think that money is for flexibility in your life, and you’d like to use it better, this is probably for you (with the bonus of your savings growing faster than you expected).