Inside the Portfolio: AI, real estate and keeping an eye on big-name IPOs

Ever wanted a deeper insight into how fund managers are seeing markets, and the first-hand thinking behind what they've been adding and removing from their portfolios?

In this new series, Inside the Portfolio, we're looking to do just that. We ask a fund manager to take us under the hood of their portfolio, explaining all the big recent changes - the ins, the outs, the overweights and underweights - as well as how they're thinking about markets more broadly.

This week, CI Global Asset Management's Geof Marshall breaks down what's changed across the CI Private Markets Growth Fund and CI Private Markets Income Fund, why the sector is watching what happens with SpaceX and OpenAI closely, and how AI is compounding the issues in private credit. 

What‘s a recent notable investment you’ve made in the fund?

There is an AI co-investment we are in the middle of closing that I will be excited to share details on shortly. This came about as a result of the relationships we have across the CI-Corient and Mubadala eco-systems. 

On a fund basis, within CI Global Private Markets Growth we made an additional allocation to Cain Global Real Estate in the first quarter. Cain pursues an opportunistic strategy of developing and investing in premium real estate assets and corporate brands that serve high and ultra-high-net-worth individuals specifically targeting premium locations in hospitality, residential, and office.

What’s a recent notable divestment or sale?

We have not yet closed out any fund positions but some of our managers have already monetised investments they have made. In venture capital, partner Radical Ventures realised an early win with the sale of Prepared911’s sale to Axon, delivering a 1.5x MOIC and triple digit IRR. 

Within private credit, Canal Road monetised its investment in Metropolis – provider of automated parking payment systems - that was purchased in July 2024 for a mid-teens unlevered IRR.

What was the most notable increase to your asset class allocation to the portfolio and why?

In lieu of the venture capital exposure in our Growth Fund, we have a strategic allocation target to royalties in our Income Fund. After months of due diligence work, we finally added exposure to CI Global Private Markets Income Fund with the Fortress Legal Assets Fund II. This strategy makes investments in legal assets, litigation finance transactions, and legal or regulatory process‑driven situations. 

The strategy targets a growing asset class that is generally uncorrelated to broader markets and supported by long‑term structural and regulatory trends. FLA II leverages the expertise of Fortress’s 30‑person dedicated Legal Assets Team, supported by the broader 670‑person Credit platform. 

Over the past 12 years, the team has invested more than $6.1 billion and realised over $5.5 billion across 106 legal assets transactions, establishing one of the most experienced and data‑rich platforms in the sector.

What was the most notable downsize in asset class allocation in the portfolio and why?

In February we adjusted our tactical allocation targets, lowering the private credit weight to the strategic target of 20%. Software was private equity’s favourite trade and largely financed by private credit. 

The AI-disintermediation risk is real. 

As Canal Road puts it, ‘the cost of software development will gravitate toward zero and undifferentiated or rules-driven workflows will be replaced by commoditised programmable logic’. Lower equity terminal values raise LTVs jeopardising refinancing risk. Yields will need to go higher and refinancing terms will tighten.

The semi-liquid private credit BDC/evergreen structure makes selling easy and it feels like we are early in this shakeout. The next leg will be selling of non-software loans to fund redemptions which will drive prices lower. There will be an interesting re-entry point when spreads widen. 

We have generally avoided software in the portfolios and have selected managers with differentiated strategies and target markets like Avenue European Special Situations and Monarch Capital Partners.

What’s been one of your most notable performers recently?

On a YTD and 1-year bases (to March 31), private equity has been the biggest contributor to returns. This should be expected, as aside from venture capital, it should be the highest returning component of CI Global Private Markets Growth Fund. Both asset classes are subject to the J-curve but I think our PE managers have been beneficiaries of two trends. 

First, being buyers in an environment where other PE funds are having trouble selling. We see 26North as directly benefitting from this trend as they have made investments in their first fund in AI ‘picks and shovels’ and media-adjacent companies. 

Second, in our view, avoiding the PE software trade went hand-in-hand with avoiding GPs reliant on financial engineering – generating returns from excess debt and multiple expansion. 

That tailwind is over. One of our GPs, American Industrial Partners, is the antithesis of this. AIP generates great returns by fixing and operating industrial businesses carved-out of bigger enterprises or otherwise neglected and un-appreciated by the financial markets.

What are the themes and trends dominating discussions right now?

In private equity, allocators have been wrestling for some time with a lack of distributions. A stubborn IPO market and a lack of interest from strategic buyers have made monetisations difficult. GPs have responded with continuation vehicles and LPs have accessed the secondaries market to rebalance, but many large institutional investors continued to make new commitments over the past few years. We are watching to see if the IPOs of SpaceX, OpenAI, and Anthropic spark new interest in IPOs.

In private credit, everyone is focused on software and how business models may be disintermediated by AI. It has long been our contention the rapid growth of private credit would lead to poor underwriting. Now the disruption of AI to software business models risks compounding those mistakes. 

In our 2026 Outlook we echoed the words of John Zito from Apollo that private credit was going to experience ‘dispersion, not distress’. More recently, speaking to poor software underwriting, Canal Road’s Mike Damasco and Don Young described it as a “vintage issue not a bubble”. 

Ultimately, this will be bigger legacy private equity problem as it is obviously junior in the capital structure to the debt. I think the lesson will be understanding the importance of an origination advantage, and appreciating the role of liquidity in portfolio construction.