Jeremy Roll has been a passive investor since 2002 and a full-time LP since 2007. Today he's invested through more than 60 active LLCs and more than 200 over the course of his career.
He got there by moving his savings out of stocks and bonds after the dot-com crash and into cash-flowing private investments. Eventually that cash flow gave him enough room to leave the corporate world. He'd spent time at Disney and Toyota. The portfolio became the job.
That history is what makes his current posture interesting. Jeremy isn't chasing every new deal. He's been deliberately defensive, holding cash and Treasuries while he waits for what he believes will be a fuller reset in real estate and public markets.
Most investing conversations are organized around what to buy. This one kept returning to a harder question: when is the correct allocation no allocation at all?
I hear versions of that question constantly from business owners and high earners. They know how to build, sell, and operate. Then liquidity arrives and cash suddenly feels like evidence that they're standing still. Jeremy's view is that this discomfort can be more dangerous than the lower yield.
Here are the parts worth your time.
The $10M problem starts after the wire clears
Jeremy has taken calls from founders who sold a business, had $10 million sitting in cash, and felt that it was "burning a hole" in their pocket. Their first concern was getting it invested. His first question was whether they had formed a view on where they were in the cycle.
Almost none had.
That landed for me because I had nearly the same conversation at a wedding last week. A founder had sold his company for roughly that amount, with more potentially coming through an earnout. He went from working 60 or 80 hours a week to a light consulting role. The capital question was tangled up with the purpose question.
When you've spent years measuring progress through activity, waiting feels irresponsible. It isn't. Sometimes the rational move is to take a breath, decide what you want the next phase to look like, and refuse to turn an identity problem into an allocation problem.
Cash is buying the option, not just the yield
Jeremy is willing to accept that someone who stays fully invested may earn more than he does. His answer is blunt: that person may do better, but he will probably sleep better.
The math isn't only today's Treasury yield against a deal's projected cash flow. It is also the value of having a dollar available when the next opportunity is materially better. A five-year private deal can become a ten-year hold. If the better deal arrives in year two, the paper return on the first allocation doesn't give you the capital back.
This is why I don't think of cash as dead money in the usual way. Cash can be expensive when it has no purpose. Cash reserved for a dislocation has a job. It preserves your ability to act when other investors are out of liquidity, conviction, or both.
AI can change the world and still be overpriced
Jeremy sees a strong parallel between current AI infrastructure spending and the internet buildout of the late 1990s. His point is not that AI fails. The internet did not fail either. The mistake was pricing short-term revenue and earnings as though demand would absorb the infrastructure immediately.
One of my closest friends owns a San Francisco construction company doing tenant-improvement work for AI businesses. The company just had its best year in more than five decades. The spending is real.
But real spending doesn't settle the return question. You can build 400 homes because the demand model says 400 buyers are coming. If 100 arrive on schedule, the community can still be valuable in the long run and the capital structure can still be wrong today.
Jeremy's forecast is his own, and he is explicit that it is a forecast. The useful part for investors is the distinction: being right about a technology is not the same as being right about the price or the timing.
Illiquidity needs to earn its premium
When Jeremy started, his rough comparison was an 8% to 10% annual public-market return against a 15% to 18% total return from the syndications he was seeing. Those were not guarantees. They were the historical assumptions behind his decision, and the gap was large enough for him to accept the lockup.
By 2016 or 2017, he no longer believed that premium was there. Private deals still carried the illiquidity. Investors just weren't being paid enough for it.
This is the part every LP needs to price explicitly. The investment isn't only an asset, an operator, and a projected IRR. It is also a surrender of control over when your capital comes home. If the expected return doesn't compensate you for that surrender, the word "alternative" isn't doing any work.
Jeremy's real estate test includes positive leverage and a return premium appropriate to the asset and the risk. Your threshold can differ. What cannot differ is the need to have one before the marketing deck arrives.
A sponsor's foreclosure follows the next deal
One of Jeremy's sharpest due-diligence points came from the last downturn. A foreclosure doesn't end when a property changes hands. It can follow a sponsor into years of higher borrowing costs on future acquisitions.
That matters to the next LP. Two sponsors can bring you similar assets with similar operating plans. If one pays more for debt because of an impaired history, you may receive a lower expected return for essentially the same property-level risk.
Fees tell a related story. Jeremy avoids structures that pay a manager primarily for assets under management. I look at the ratio: what does the sponsor earn if a deal reaches 80% of pro forma, and what changes if it reaches 120%? A manager needs enough recurring revenue to build a competent team and operate through a difficult decade. But financial stability should not turn into prosperity that is indifferent to investor outcomes.
Alignment is not "low fees." It is a compensation structure that keeps the operator engaged and makes the meaningful upside depend on performance.
The best time will feel like the worst time
Jeremy's line at the end of the real estate discussion was the episode in one sentence: the next window may be the best time to invest and the hardest time to invest at exactly the same time.
Investors have spent years dealing with frozen transactions, challenged sponsors, floating-rate debt, capital calls, and delayed exits. Some have left the space completely. That response is understandable. It also creates the conditions in which patient capital can have better basis, better terms, and more leverage with operators who still want to buy.
For someone starting now, Jeremy's advice is to use the waiting period. Learn the cycle. Form your own view. Start with one asset class you can understand, then carry the underlying business lessons into the next one.
Most importantly, don't confuse speed with competence. Real estate reprices slowly. Jeremy watched investors catch falling knives in 2008 and 2009, years before many assets found their trough. His preference is to be late rather than early.
I think that is the right final frame. You do not need to call the bottom. You need to preserve enough liquidity, judgment, and emotional room to recognize a good risk-reward when it finally appears.
Give the full episode a listen. Jeremy has spent nearly 25 years living with the consequences of LP decisions, and his case for patience is more useful than another list of deals to chase.
- Sam Silverman
Silverman Capital
