We've ended up in a world where nearly every rebalancing tool calls itself “automated”, regardless of how automated it really is.
Words wear out. “Automated.” “At scale.” “Efficiency.” These terms are used so often and so loosely that they begin to lose their meaning. So we end up in a world where nearly every rebalancing tool calls itself “automated”, regardless of how automated it really is.
At Smartleaf, we use the analogy of a robot versus a power screwdriver. There’s nothing wrong with a power screwdriver. If you’ve ever tried to assemble a deck with a hand tool, you know how much faster the powered version is. But a power screwdriver still requires you to place every screw, hold the boards, and do the work. A robot, by contrast, doesn’t just help you place screws—it assembles the deck for you.
Smartleaf is a robot. Many other systems—very capable in their own right—are power screwdrivers. They provide alerts and suggestions. They make it easier to do the work you were already doing manually. That’s valuable. But it’s different from what Smartleaf does.
If we’re going to talk honestly about rebalancing technology—what it means, what it can do, and how to use it—we need to reclaim the word “automated.” Not as a marketing slogan, but as a precise description of a workflow where the machine doesn’t just assist advisors—it, well, automates entire tasks, freeing advisors to actually advise.
The Confusion: When “Automated” Means “A Little Easier”
I’ll begin with a story. It’s a conversation I’ve had dozens of times with firms evaluating Smartleaf:
Q: How many staff would it take to rebalance 100 accounts?
A: Two people.
Q: What about 1,000 accounts?
A: Two people.
Q: 10,000?
A: Still two.
By the time we get to 100,000 and the answer is still “two,” people understandably raise their eyebrows. They ask: “But what about tax management?” “Custom models?” “Social screens?” “Cash-out requests?” “Transitions?” And each time the answer is the same:
There is no difference. The system handles all of it automatically.
This is the moment when the idea of automation—in the true, literal sense—starts to sink in. It’s also the moment you can feel the fog lift a bit. Because there’s an unspoken assumption baked into many tools that call themselves “automated”: that humans still need to make decisions one account at a time, one trade at a time. The system may highlight where attention is needed, but the human still holds the screwdriver.
If your system produces alerts (“Hey, you may have losses to harvest”), that’s helpful. But it’s not the same thing as robot-driven rebalancing. Helpful suggestions are not automation. They don’t replace human labor; they simply inform it.
True Automation
Automation, real automation, is more than existing workflows made incrementally more efficient. It’s a change in how portfolio management is done. With full automation, the system:
Tax preferences, screens, asset allocations, constraints, risk targets, and product choices are no longer scribbled on sticky notes or relegated to the depths of advisors’ memories. They’re codified in parameters. Every portfolio carries with it a machine-readable description of how it should be managed.
The system doesn’t react to one-off triggers (“it’s tax-loss harvesting day!”) or a patchwork of alerts. It performs a holistic, portfolio-level evaluation: Considering drift, taxes, costs, and client instructions, which trades would improve this portfolio right now?
There’s no “review this suggestion and decide what to do.” The system generates trading instructions that are ready to send for execution.
Automation doesn’t mean no human involvement, but the reasons an automated system might need human help are rarely what people expect. It isn’t tax complexity or customization or tricky combinations of cash withdrawals combined with tax loss harvesting, asset allocation changes and new security restrictions. Those are easy to automate. The real sticking points are mostly mundane data issues, a problem of garbage in, garbage out: late corporate actions, unknown securities, maybe clients who are unsure of their preferences.
“Full” automation still includes a place for humans—just not in deciding whether to harvest a loss or swap a fund. Humans intervene when the system can’t know what to do because the data is wrong or incomplete, or because something unusual is happening. (e.g., “Don’t trade this account until next week; the client is transferring in cash”) That’s it. Humans aren’t there to complete the incremental tasks; they’re there to fix what prevents the system from completing the incremental tasks.
The Automated Rebalancing Workflow—What It Looks Like in Real Life
When fully automated rebalancing is operating as intended, the workflow is simple and repeatable:
1. Suspend the accounts with bad data
Flag them, fix the data, move on.
2. Trade accounts with mandated actions
Cash withdrawals, compliance requirements, “never own X,” model changes—these trades are auto-generated.
3. Trade accounts with high cost/benefit scores
This is where optimization shines. Every account is analyzed, every day, for the net benefit of trading—benefits minus costs. If the net benefits exceed a threshold, trades are generated.
Some firms add variations (different thresholds for different groups of accounts, calendar-driven overlays, review of outliers), but the essential workflow is the same. And crucially, none of these choices require more human labor. They only alter the details of what robot does, not the level of automation.
A Different Way to Work: Roles Become More Specialized
One of the perhaps surprising implications of true automation is that it changes roles and responsibilities across the firm:
Advisors are no longer traders. They design each portfolio—expressed through parameters like risk targets, screens, asset class adjustments—and then the system handles the rest. Advisors spend more time with clients because there’s no need to check whether each client is “in alignment.” They always are.
Investment policy committees define strategy. Then it gets translated into models and rules. Every portfolio reflects that intellectual capital automatically.
Overlay teams oversee workflows. They handle data issues and other anomalies. They execute but do not create trade tickets.
Menu-Driven Customization: The Quiet Revolution
“Automated” is sometimes equated with “cookie cutter”. It’s efficient, but you give up customization. This is backwards. With portfolio management, automation mean more, not less, customization. Because when customization is built into the system as parameters—not manual choices—adding a client preference is as simple as checking a box:
There’s no incremental labor. It becomes free to deliver. So firms stop worrying about whether they can “afford” to offer personalization.
Holistic Cost/Benefit Analysis (or, Every Day Is ‘Improve the Portfolio’ Day)
Traditional workflows tend to be action-driven:
Each action is isolated and can conflict with another. Should you harvest a loss if it creates additional drifts? Should you rebalance now or wait for a cash distribution? Should you incorporate that model change or delay it because of taxes?
These decisions involve trade offs. Humans trying to answer these questions one tradeoff at a time, one account at a time, quickly become overwhelmed.
A holistic system solves this by asking a single question daily: “Which set of trades improves this portfolio the most right now, given everything we know?”
That’s the robot.
Why This Matters (And Why It Isn’t About Bragging Rights)
Automation—real automation—matters because:
Smartleaf isn’t magic. But it is special. It’s a system built from the ground up to let the robot do the work. Other tools do valuable things. They genuinely help advisors who want a better screwdriver.
But we shouldn’t confuse “making it easier to do the work” with “doing the work for you.”
In the End, It’s Not About Automation—It’s About Time
If the robot can do the trading—not just help you trade, not just suggest trades, but actually do the trading—then you get to spend your day doing something no robot can do: talking to clients, understanding their lives, and delivering advice that is uniquely human.
And that’s the point. That’s the true value of automation.