Why I Track Cashflow and Balances Separately

Double-entry accounting systems are powerful tools for finances, but it's a bit overkill for personal finances and leads to more friction to inputting data. Balancing the books is tedious and time-consuming and offers very little benefit when you're not held to exacting accounting standards.

For instance, I don't track my gross salary in cash flow – only my take-home pay that lands in my checking accounts. The portion of my salary that ends up in retirement accounts I trust will show up as an increase in my balance sheet.

I'm too lazy to hunt down the exact amount of contributions that get deducted from my salary every other week.

Strategic laziness is a benefit. I can rely on my balance sheet to capture the broad picture, and I use my checking accounts' cashflow to dig into the details of monthly life. Most crucially, my time is valued by skipping over detailed account tracking.

"Well wait," you may say. "If you value your time, why not automate your finances?"

The short answer? Automation bias and related phenomenon.

The Long Answer:

Repetition breeds familiarity, and taking the time to log into each account every month is worth my attention.

I like keeping my hands on my accounts every month to be sure I'm comfortable with them and know what's going on. I take the opportunity to read any platform-specific alerts or messages.

My approach is not for everyone. There's value in certain types of automated finance tools, and I hear good things about net worth trackers like Personal Capital.

Instead my approach is about finding my personal sweet spot between fully automated and fully manual.

On one extreme of having everything manually inputted, you waste a lot of time fiddling with error-prone data entry and report generation. You'll want some automation to remove repetitive tasks like building reports, backing up records, and validating inputs.

Input validation is an especially important step. For instance with my finance app, I had it so that you could only input expenses that happened within a couple months. This helped reduce the odds of mistyped dates, especially when the year was off by one.

On the other extreme with total automation, you mindlessly connect up your applications and don't need to think too hard about what's actually going on. You may glance over automated reports without digging into the details. Worse still, I've seen automation keep running whilst people ignore its output entirely.

They take things for granted if they don't have to work on them directly.

You're also subject to garbage in, garbage out. I love my personal system for categorising expenses, but nobody else uses the exact same categories as I do. Also in all my [limited] experience, no financial institution has an accurate system for reporting the category of all expenses.

Sure you might get an institution that labels credit card transactions with rough groups like "dining out" or "household" or "health." Yet those labels are not accurate enough to rely on, and you still need human oversight to verify them. Sometimes you also get expenses that slip through the cracks and are unlabelled.

This is one area where classic automation bias comes in. A person overseeing the correct labelling of expenses may second guess themselves, or blindly trust the label, or get exhausted and stop trying to fix broken labels.

They end up fighting the automation, instead of benefiting from it.

There are just certain tasks that are better left without automation getting in the way – those that require human judgement. Instead of trying to accurately predict the judgement, automation should focus on making it easier for the human to finalise their decision.

Automate the Good, Discourage the Bad:

You want to make good things automatic and bad things difficult to do. Barriers and incentives are a great way for controlling my dearheart without him noticing. They're also a key factor to finding the sweet spot to automation.

For instance, by requiring that he manually input every expense each month my dearheart feels the pressure to make less transactions every month. A move toward frugality for someone with a history of voluminous spending. All because he knows that if he buys something, he'll have to type it in later – and his laziness might trounce his impulsivity.

Stumbling Blocks and Edge Cases:

Another reason I avoid heavy automation is because of a phenomenon I dub the stumbling block: it's where you get 90% of the way through setting everything up and then discover that the remaining 10% is impossible to complete.

For instance, you go ahead and connect almost all of your credit unions and brokerages to an app, then discover that your last account doesn't offer a connection.

What do you do?

Well, either you accept the inaccuracy of missing an account, or you fall back to more primitive but reliable methods for a centralised report.

Whilst all my accounts right now may integrate cleanly with apps, there's no guarantee that the integration will be supported in the future. Nor is there a guarantee that I will not open a new account that doesn't integrate.

Opinionated Philosophy To Do One Thing Well:

As mentioned at the start of this article, double-entry bookkeeping is tedious overkill. Likewise, a monolithic application that does both asset tracking and cash flow review is more likely to be mediocre at both.

See, I have strong opinions about how my finances should be run.

One of the philosophies I sometimes subscribe to is that tools should do one thing, and do it well. That it's better to use multiple tools together than one single large tool.

In my case, "doing it well" is intrinsically tied up in my opinions about finances. I believe in stuff like temporal context – a fancy way of saying that information should be presented in a way that makes use of time. For instance, comparing several months at once, so you're not just looking at numbers in isolation.

I also believe that net worth is a vanity metric, unsuitable for serious tracking. Instead I prefer to focus on underlying systems and processes that grow my net worth as a byproduct. Tracking those is more important than seeing my net worth.

Closing Thoughts:

I want my accounting to be streamlined to input and review, but not so smooth that it takes no thought.

I want it to be quick, but not painless. There should be enough attention-grabbing work that I get intimately familiar with the accounts I'm tracking.

My finances are the pulse at the core of my household. Keeping my fingers steadily on that pulse is key to our eventual success.

Thus I use my balance sheets for the big picture, and my cash flow reviews for the minutiae. They each accomplish separate goals, which allows me to tweak those goals without affecting each other – something a monolithic tracking approach would struggle with.