Does your 401(k) balance stack up against the new national average?
Is how America is saving for retirement in line with how you are saving? Vanguard's 'How America Saves' report on the state of 401(k) participation is now in its 24th year, and for the moment, all the trend lines are pointing up. That's good news for the most part, but there are some cracks beneath the surface. While 401(k) savings rates and balances are up, so are hardship withdrawals.
First, the positive: The average 401(k) balance among nearly 5 million accounts in Vanguard's data set was $148,153, up over 9% from the year before. The average participant rate of return was 13.7%. The average employee percentage of salary contributed, known as the deferral rate, was 7.7%, and when combined with employer contributions, it was 12%.
My job is offering me a payout. Should I take a $61,000 lump sum or $355 a month for life?
'He doesn't seem to care': My secretive father, 81, added my name to a bank account. What about my mom?
Israel-Iran clash delivers a fresh shock to investors. History suggests this is the move to make.
I'm 75 and have a reverse mortgage. Should I pay it off with my $200K savings — and live off Social Security instead?
Coinbase's stock sees a 'golden cross.' Why it may not be a bullish signal to buy.
Perhaps the best news of all is that the proportion of participants who increased their contribution rate rose to 45%, with 29% automatically increasing and 16% taking active steps to increase.
'One might expect to lose ground in tough years, but we're seeing one positive year following the other,' said David Stinnett, head of strategic retirement consulting at Vanguard. 'That's attributed to automatic plan designs that provide this suit of armor around the 401(k) system. It's a very positive thing to see.'
Despite the rosy general picture, when you dig down deeper, there are some issues to be concerned about for retirement savers.
The 'average balance' measure among so many participants can be deceptive because it includes many so-called 'super savers,' who max out all the available options to them, plus older and higher-paid employees. If you measure by the median, which is looking at the middle of the pack, the numbers are lower. Median 401(k) balances were only $38,176 in 2024, with a participant contribution rate of 6.8% and a combined rate with matching of 11.5%. During the year, 8% decreased their contributions and 2% stopped them altogether.
Another area of concern is withdrawals. Some 80% of plans allow loans, and consistently in the past five years, 13% of participants have had an outstanding loan balance. Withdrawals have been going up in recent years, however. Some 4.8% of participants took hardship withdrawals and 4.5% took non-hardship withdrawals in 2024, up from 1.7% and 3.4% in 2020, respectively. That reflects both worsening economic conditions and loosening restrictions on withdrawals.
It also may reflect the degree to which the automatic plan enrollments are working, because plans now include many more low-income workers than previously. 'It's not necessarily bad,' said Stinnett. 'This includes many lower-income workers who would have been experiencing hardship before, but they weren't in the data because they weren't in the plan before.'
For the most part, 401(k) participants stick with the hand they are dealt. Almost 90% of plans offer an automatic deferral into a managed target-date fund pegged to the participant's age, and most just stick with that. While the average number of investment choices that plans offer is 17.5, people make an average of just 2.3 choices, and 59% are using a single target-date fund.
Only 7% of the 5 million participants surveyed were exhibiting what Vanguard called 'extreme participant asset allocations,' which means they were invested 100% in either equities or fixed-income options. Only 5% took part in any sort of trading during the year. 'Trading activity in 2024 remained at the lowest level in nearly two decades,' Vanguard noted.
Women tended to be more conservative and less active than men overall when it came to 401(k) investing. They traded less frequently, Vanguard found, and were more likely than men to hold a single target-date fund.
This is not a bad data point, Stinnett said. 'We see some very positive things across the board as it relates to women investors: strong saving behavior, saving at higher levels, and yes, from an investing standpoint, they tend to invest in target-date funds and tend to trade less. But the target-date fund offers such a great solution for people because it covers all the bases.'
Among super savers, only 14% maxed out their 401(k) contributions to the statutory limit of $23,000 in 2024. Only 16% of those eligible made catch-up contributions, which allowed people 50 and older to contribute an additional $7,500 in 2024. Starting in 2025, those between 60 and 63 can contribute up to an additional $11,250, but given these numbers, utilization is likely to be low.
Only 18% of participants used the post-tax Roth 401(k) savings feature now offered by most plans. And only 10% used the after-tax contribution feature, which allows participants to make extra contributions above the $23,000 limit and then usually convert them immediately to Roth funds in a backdoor conversion move.
'The takeaways for me are positive,' said Stinnett. '2024 was a very strong year for equity markets, and that's an important tailwind. But we're already seeing similar good behavior halfway through this year, and it's been choppy in the markets.'
Got a question about investing, how it fits into your overall financial plan and what strategies can help you make the most out of your money? You can write to me at . Please put 'Fix My Portfolio' in the subject line.
You can also join the Retirement conversation in our .
More Fix My Portfolio:
Two ETFs that have beaten value stock indexes with this simple approach
S&P 500 nears record as stocks soar, oil sinks and investors throw caution to the wind
We're living in 'end times' when you can't retire on $1 million
Why this banking proposal may mean good news for the bond market and investors
My cousin died before claiming his late father's $2 million estate. Will I be next in line?
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Bloomberg
26 minutes ago
- Bloomberg
Fed's Daly Says Muted Tariff Impact May Open Door to Cut in Fall
By Catarina Saraiva and Updated on Save Federal Reserve Bank of San Francisco President Mary Daly said she's seeing increasing evidence that tariffs may not lead to a large or sustained inflation surge, helping bolster the case for a rate cut in the fall. 'My modal outlook has been for some time that we would begin to be able to adjust the rates in the fall, and I haven't really changed that view,' Daly said Thursday in an interview on Bloomberg Television.
Forbes
26 minutes ago
- Forbes
AI Gave The World Infinite Content—Now What?
Tejas Manohar is the cofounder/co-CEO of Hightouch. Just a few years ago, generative AI (GenAI) felt more like a curiosity than a tool. We asked language models to write love letters in the style of tech bros or explain quantum physics to a 5-year-old. Visual platforms responded to prompts like "a dragon in a business suit, pixel art style" or "a Renaissance portrait of a barista." The results, while novel and amusing, were rarely practical for business. That has changed. By the end of 2024, GenAI outputs became sharper, more polished and increasingly indistinguishable from human-created work. In 2025, with tools like GPT-4, Midjourney, Runway and Canva AI becoming widely adopted, content creation is no longer the bottleneck it once was. Soon, marketing teams will be able to generate dozens of creative options in minutes. However, this shift introduces a new problem: With so much content, how do we decide what to use, for whom and when? Most marketers are now using GenAI to create assets. While Salesforce reports that 76% of marketers use AI to generate content, the processes for deploying that content haven't evolved. The typical workflow still involves pasting AI-generated copy into spreadsheets, testing a couple of variants, manually picking a winner and repeating it all. That might work in the short term, but it's not scalable. More importantly, it doesn't improve over time. More content is not the solution unless there's a system to decide which content to use and how. Imagine an orchestra where every musician trained at Juilliard, but there's no conductor. That's what marketing looks like in a GenAI world without decisioning. There's creativity, but no coordination. Marketers today face a flood of assets, but the bigger challenge is figuring out what to send, to which audience and when. These are not creation problems. These are decisioning problems. And we're still trying to solve them using tools and mental models—journey builders, marketing calendars and simple A/B tests—built for a world where content is scarce. Traditional workflows assume that you'll create a handful of subject lines, define a few segments and test some variations. But GenAI doesn't create one or two options—it creates hundreds. Suddenly, you're staring at thousands of possible combinations across messaging, timing, audience and channels. Marketers can't test every option. They can't manually orchestrate every journey, and they certainly can't rely on batch-and-blast methods anymore. A new approach is needed. For many organizations, AI decisioning has become a key part of their AI strategy. This new category of technology sits between content creation and content delivery. It enables marketers to deploy AI agents that make real-time decisions about which content to send to which user. These systems use reinforcement learning (the same type of machine learning behind self-driving cars and streaming recommendation engines) to optimize for business outcomes like conversions, retention or lifetime value. Think of how platforms like Google and Meta Ads operate. You set your goals, upload creative assets and the system optimizes combinations to deliver results. Now imagine that same model applied to email, push, in-app messaging and CRM. That's what AI decisioning aims to achieve, only this time with transparency and control built in. To adopt AI decisioning effectively, companies need to get the basics right first. That means clarifying goals, improving data access and identifying where manual decisions slow things down. Start small by pinpointing bottlenecks in your workflow, whether that's testing content, segmenting audiences or managing channels. Silos are a major hurdle. When teams like marketing, data and product work in isolation, decisioning falls flat. Aligning around shared goals, metrics and timelines helps break down these walls and ensures AI systems have the inputs they need to be effective. The best way to begin is with a focused use case, such as optimizing subject lines or send times. Prove value quickly and then scale. AI decisioning is not about replacing everything at once; it is about creating a system that learns and improves over time. Used together, these technologies form a closed-loop system. GenAI generates content while AI decisioning systems select the right assets for each user based on performance data. As results come in, those insights feed back into the content generation process, allowing both creation and decisioning to improve continuously. GenAI acts as the input layer, creating at scale. AI decisioning functions as the optimization layer, learning what works and when. Combined, they create a flywheel where content fuels decisions and decisions enhance future content. But none of this works without human oversight. Marketers still need to be involved. AI systems must be transparent, auditable and accountable. Teams need to know how decisions are made, what experiments are running and have the ability to approve content and manage risks. In the coming months, content bottlenecks will fade as GenAI becomes even more integrated into daily workflows. But that's only the first step. The true differentiator will be how effectively teams can deploy the content they generate to drive meaningful results. The winners in the next era of marketing won't be the ones who generate the most creative assets. They'll be the ones who build systems that know what to do with them and can adapt in real time. So keep prompting and creating. But remember: the next meaningful shift in marketing won't just come from creation—it will come from smarter decisioning. Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Bloomberg
28 minutes ago
- Bloomberg
S&P 500 Homes In on Record as Trump Posits Powell Replacement
US stocks resumed their climb Thursday with the S&P 500 Index narrowing in on a record following a report President Donald Trump will announce his pick for Federal Reserve Chair Jerome Powell's replacement earlier than anticipated. The S&P 500 rose 0.4% at 9:32 a.m. in New York, placing it within spitting distance of an all-time high. The technology-heavy Nasdaq 100 Index bounced roughly the same amount, extending its record run. The Cboe Volatility Index, or VIX, fell below 17 even as Trump's self-imposed tariff deadline looms alongside uncertainty around whether peace can hold in the Middle East and souring economic data.
