401k-allocator

401k Allocation Optimizer — How to Use
This tool helps you quickly translate our firm’s model allocation into a specific fund mix using whatever investment options the client’s 401k plan offers. You enter the funds available in the plan, set the U.S. equity portion, and the tool produces a recommended weight for each fund along with a CSV you can drop into our reporting software.
Step 1 — Enter the funds available in the client’s 401k.
Click “+ Add fund” and enter each U.S. equity option from the plan. You’ll need the ticker (or whatever identifier the plan uses), a fund name, and the 9-cell Morningstar style box.
To get the style box numbers, find the fund on Morningstar.com and look at the Portfolio section — there’s a 3×3 grid showing what percentage of the fund’s holdings fall into each style category. Copy those 9 numbers in this order: Large Value, Large Blend, Large Growth, then Mid Value/Blend/Growth, then Small Value/Blend/Growth. Paste them into the “Paste 9 numbers from Morningstar” field and the grid below will auto-fill. The numbers should sum to about 100.
Repeat for each U.S. equity fund offered in the plan. You don’t need to enter bond funds, international funds, target-date funds, or money market funds — this tool only handles the U.S. equity sleeve.
Your fund list saves automatically in your browser, so the next time you open the tool, your previous funds are still there.
Step 2 — Set the U.S. allocation percent.
This is the percentage of the client’s total portfolio that should be in U.S. stocks based on our model. For example, if the model is 60% U.S. equity, 25% international, and 15% bonds, enter 60. The recommended fund weights will be scaled to sum to this number, leaving room for the international and bond allocations the client will set up separately.
If you forget to update this between clients, hit Reset on the target allocation section to bring the firm model back, and double-check the U.S. allocation field for the new client.
Step 3 — Confirm the target allocation.
This pre-fills with our firm’s model U.S. equity breakdown. You usually won’t need to change anything, but you can edit any cell if a particular client situation calls for a tilt. The total must equal 100% before the tool will optimize — if you change the values and they don’t sum to 100, you’ll see an error when you click Optimize.
Step 4 — Click “Optimize allocation.”
The tool finds the combination of available funds that comes closest to matching the firm model, given what the plan offers.
Step 5 — Review the results.
You’ll see two things:
The Allocation table shows each recommended fund and its weight as a percent of the client’s total portfolio. These are the numbers the client (or you, on their behalf) will enter on the 401k website.
The Target vs achieved table shows how closely the recommended mix matches our model. Most plans can hit it within a couple of percentage points per category. If a category shows a meaningful gap, you’ll see a note at the top of the results explaining what the menu couldn’t reach. This is normal — many plans don’t have dedicated small-cap value funds, for example. Use your judgment on whether the gap is acceptable or whether a rollover conversation makes sense.
Step 6 — Export the CSV.
Click “Export CSV” to download Allocation.csv. The file has two columns — Holdings identifier and Weight — and is formatted for direct import into Vanguard’s portfolio analytics tool.

A few practical notes
The tool saves your fund list, target allocation, and U.S. allocation percent in your browser. When you start a new client, you’ll usually want to either remove the previous client’s funds (using the Remove link on each row) or just edit them in place if the new client’s plan has overlap with the previous one.
If two advisors use the tool from different computers, they each have their own saved state — there’s no shared database.
The tool is for U.S. equity only. International and bond allocations are not handled here; those are part of our separate model documentation.
This is an internal tool, not client-facing. The output is a starting point for our recommendation, not a final document. Apply judgment.

How the optimizer works
The tool takes two inputs: a list of available funds, each with a 9-cell Morningstar style box describing how the fund’s holdings break down across large/mid/small-cap and value/blend/growth, and a target allocation expressing how you’d like your overall portfolio distributed across those same 9 categories.
The tool then searches for the combination of fund weights that, when blended together, produces an overall portfolio whose category exposures come as close as possible to the target. “As close as possible” is measured by sum of squared differences across the 9 categories — meaning a 5-percentage-point miss in one category counts the same as roughly 3.5 points of miss spread across two categories, and large misses are penalized disproportionately more than small ones. This is the same standard “least squares” criterion used in regression analysis.

Two constraints are enforced during the search: every fund weight must be zero or positive (no shorting), and the weights must sum to 100% (fully invested). The search algorithm is called projected gradient descent — at each step it nudges the weights in the direction that most reduces the mismatch, then projects the result back onto the set of valid allocations that satisfy both constraints. This repeats until the weights stop changing meaningfully, which typically happens in well under a second.

The continuous weights produced by the search are then rounded to whole percentages using the largest-remainder method, which guarantees the final allocation still sums to exactly 100%.

What the optimizer is and isn’t doing
It’s solving a pure style-matching problem. It does not consider expense ratios, historical returns, fund quality, manager tenure, tax efficiency, or any forward-looking expectation about which categories will outperform. It assumes the style box numbers you’ve entered accurately describe the fund and that those characteristics are stable over time — neither of which is strictly true, since funds drift and Morningstar updates its classifications periodically.
When the available funds can’t reach the target — for example, if you want exposure to small-cap value but no fund in the menu holds small-cap value stocks — the optimizer will still produce its best fit and flag the gap rather than refuse to answer. The “Target vs achieved” comparison is the honest record of what the math could and couldn’t do.