If you’re reading Retirement Upside, you’ve probably heard of the 4% rule.

In case you haven’t, it’s generally considered a safe income planning guideline that lets retirees withdraw 4% of their total savings in the first year of retirement and then adjust that dollar amount for inflation every year afterward to make their nest egg last for 30 years. The rule is ubiquitous in the mainstream media and among popular financial planning voices online, thanks to its straightforward answer to a daunting financial planning question: How much can clients spend in retirement without risking poverty in old age? But they may be surprised to learn that the rule’s creator, Bill Bengen, never intended it to be a baseline for retirement income planning. Quite the opposite, in fact.

“It’s actually a rule that only applies for a very narrow segment of the population in practice,” Bengen recently told Retirement Upside. Ultra-risk-averse people living off a 401(k) account who want to make sure their retirement income plan would have worked out without adjustment during the worst period in the modern history of the stock market should follow it. Others who are accepting of more risk can afford to draw more income, especially if they’re willing to make adjustments along the way.

“I never intended it to be a panacea for ‘safe’ retirement income planning, but that’s kind of what it’s become,” Bengen said.

Keeping Busy

Anyone who has followed Bengen’s work in recent years will already know the 4% rule isn’t the end of the story, as the retired financial advisor and longtime retirement researcher has kept busy since leaving professional practice behind in 2013. Not only has he reassessed the “safe” inflation-adjusted spending rate, raising it to 4.5% in 2006 and to 4.7% in 2021, but he has also looked closely at the emerging category of risk-based guardrails and other income-planning frameworks. His latest analysis, for example, reviews what is known in academic circles as the Guyton-Klinger (G-K) Decision Rules withdrawal strategy. That sounds complicated, but its basic principle is to begin with a more aggressive income amount and then annually adjust retirement withdrawals up or down (or not at all) based on changes in the current withdrawal rate and portfolio performance.

“It attempts to mitigate the risks of a pure cost-of-living adjustment scheme,” Bengen said. These include the potential for significant underspending early in retirement, as well as the risk of lower-than-necessary cumulative spending when portfolios overperform. “It’s potentially more appealing in practice because most Americans are comfortable taking some risk with their retirement income plan and making adjustments along the way.”

Running the Numbers

Bengen has the benefit of being personal friends with Jonathan Guyton, one of the researchers for whom the flexible spending technique is named. “He helped me understand the intricacies of G-K,” Bengen said. “Despite his valued assistance, it took me almost two months to develop a process flow chart and prep my Excel spreadsheets to do the analysis. That’s why you haven’t heard from me in a while.”

So what did Bengen’s latest work find? Put simply, retirees who are willing to follow the G-K method actually have an average maximum safe starting withdrawal rate of 9.6%, which is far higher than anything he’s encountered with a conventional cost-of-living adjustment strategy. There’s a catch, though.

“G-K does a superb job of producing high initial withdrawal rates, which are amazingly consistent across all market environments,” Bengen said. Its shortcoming is that the cumulative real withdrawals, or the sum of the real value of all withdrawals made during the entire 35-year planning horizon, are generally lower. In fact, for all 361 retirement scenarios examined, save one, the conventional inflation-adjustment strategy generates superior cumulative real withdrawals, frequently by a large amount. This is an important fact for financial advisors to hear, he said, because the G-K approach is growing in popularity in the industry.

“It’s useful for some in the population, including people who agree to the idea of higher income early,” Bengen said. “That could work for people, but you have to understand that the reduction in cumulative income could be 60% or 70% in extreme scenarios. That’s a risk to consider.”

On to the Next Challenge

What’s next on the docket for Bengen? He’s ready to take on Warren Buffett’s retirement income advice.

Berkshire Hathaway’s 2013 shareholder letter says Buffett instructed the trustee for his wife’s bequest to put 10% in short-term government bonds and 90% in a low-cost S&P 500 index fund. During good times in the market, income can be drawn from the portfolio, but the cash reserves should fund spending needs during bear markets when stock prices have fallen.

“That approach makes sense on first look, but you have to test every idea thoroughly when it comes to retirement income planning,” Bengen said. “Things that seem to make total sense need to be tested, and his proposal is quite a bit different from other schemes. I should be done with the analysis in about a month. It takes a while to set up my spreadsheets and procedures.”

Remember 2004? It was the year that Facebook officially launched, Shrek 2 dominated the box office and, of course, Janet Jackson endured the most publicized wardrobe malfunction of all time. It was also the last time the iShares 20+ Year Treasury Bond ETF (TLT) traded this low.

Rising yields on long-term Treasuries pushed the price of the largest long-term bond fund to its lowest price in 22 years. TLT fell below $82 per share on Friday as risk-averse investors fled the asset. Higher yields on long-term bonds have been a steady trend over the past four years, and currently stem from scarcity caused by high demand for capital to fund the artificial intelligence buildout, as well as higher inflation expectations due to the war in Iran.

“Investors are likely risk averse to owning longer-term Treasury debt right now,” said Jason England, a fixed income strategist at Simplify Asset Management. “Most people want to favor the front end or intermediate part of the duration curve right now, if you’re going to take any duration on at all.”

Short Over Long

The State Street SPDR Bloomberg 1-3 Month T-Bill ETF (BIL) has outpaced TLT over two-, three-, five- and even year-to-date periods, said Athanasios Psarofagis, an ETF analyst at Bloomberg Intelligence. It’s a sharp reversal from TLT’s heyday: Before 2022, when the inverse relationship between stocks and bonds still held, the fund thrived as a portfolio diversifier. “That was the golden era for TLT,” he said. “Some of the best years might be behind it.”

Here are some more numbers on the fund’s decline:

• TLT has fallen more than 50% from its high of $171.11 in 2020.
• The fund’s assets peaked at $64.5 billion, but have since fallen back to $41.6 billion, according to ETF.com data.

Rate Expectations: Long-term bonds can have wild price swings, and traders have historically tried to bet on timing TLT to benefit from price increases after rate cuts. “If they really think rates are going to move a certain way, it’s almost like standing at the tail end of a seesaw,” said Psarofagis. “If someone jumps on it, you’re getting this giant move up.” Except everyone usually gets it wrong, he said. “It’s been something that has literally not worked for four years, so it’s still kind of baffling that it has the amount of assets that it does.”

What’s a bigger myth: The Odyssey or a hedge fund that gets every call right? In the case of Leopold Aschenbrenner’s AI-focused Situational Awareness, overleveraging on the biggest trade of the past two years proved there are plenty of funds that guess wrong.

The firm was pushed to the edge of implosion last month before offloading most of its public equities portfolio to Ken Griffin’s $71 billion Citadel. Griffin’s firm came out on the right side of things, instantly converting the deal into one of its best months in years. Broader markets have, so far, benefited also.

Rally Around Relief

By late July, highly leveraged Situational, which at one point was up 400% this year, was feeling the sting of a sharp, monthlong selloff in AI-adjacent equities. South Korean memory chip manufacturer SK Hynix and cloud computing firm CoreWeave, two key holds, traded 50% below their peak at points. The tech-weighted Nasdaq-100, home to other important holdings, officially slipped into correction territory. Then, on July 29, Citadel approached Situational about offloading its distressed equities. In less than 24 hours, Griffin’s firm acquired the lion’s share of Situational’s $16 billion holdings in public companies at a 10% discount.

The deal proved a near-immediate coup for Citadel, according to multiple reports. Its flagship Wellington Fund was roughly flat in July before the deal but closed the month up 5.9%. That made for the best month since 2022, and Wellington is now up 12% in 2026. On top of that, Citadel’s tactical trading fund gained roughly 11% in July and its equities fund roughly 14%, a record advance for both.

Griffin’s gain may have provided just what the market needed in a moment of wavering confidence. Situational’s exposure during the AI stock rout was exacerbated by a perfect storm of heavy borrowing and short bets on traditional software equities, turning the fund into a forced seller. Citadel’s intervention put a stop to the $24 billion fund’s unraveling, substituting a more stable investor in no hurry to dump assets:

• In the past week, Citadel’s acquisition has helped to power a significant relief rally. Several stocks that formerly made up Situational’s core positions have proven incredibly resilient in the last five trading sessions: CoreWeave is up 47%, SK Hynix 19% and SanDisk 33%.
• The hyperscalers financing the AI buildout have also helped soothe market jitters about the trade. Amazon, Alphabet, Meta and Microsoft have reaffirmed their plans for massive AI-related capital expenditures in 2026 and beyond, much to the benefit of AI-adjacent companies.

Out of Ruins, an Empire: Griffin’s fund raided Enron for its top talent after the energy company collapsed in 2001 and bought the books of failed competitors Amaranth Advisors and Sowood Capital later in the decade. One could say, when it comes to Citadel, that the firm has mastered the alchemy of pulling the proverbial phoenix from the ashes.