Summary: Vgi Partners Global Investments (VG1) launched its $550m LIC in Sep 2017. In part this was off the back of some high profile short-selling battles: Slater and Gordon and Corporate Travel Management. Long-short funds sound attractive to many retail investors who consider themselves to be getting access to greater sophistication and better risk-adjusted returns.

In reality, long-short funds and hedge funds have terrible performance (average and median) after expenses. This is largely due to their inflated Total Expense Ratios, where high costs and fees detract from returns. Typically, investors are paying fund managers to make different types of bets (sectors, asset classes, timing, short selling, currency), but the outcome is heads: they take most of the winnings; tails: you take all of the losses. Nevertheless, retail investors eagerly jumped into VG1's $550m IPO and 2019 $300m capital raising.

VG1's performance started ok (and it traded at a premium), but performance then quickly fell away, and in the subsequent six years, it has significantly underperformed. Overall, from 30 Sep 2017 to 19 Oct 2023, VG1 has delivered an appalling Total Shareholder Return of -0.34% annualised compared to 11.48% for VGS.

Since 2020 there have been multiple attempts to resolve VG1's discount via buybacks, increasing franked dividends, a merger of VGI Partners and Regal, and activism to wind-up or convert to an open end structure. And, once again, VG1 is in the news with Saba Capital (a Closed End Fund discount activist) quickly building a 6% stake.

In this post, I focus on how enormous the holding costs can be of trying to capitalise on ASX CEF discounts, or the potential for them to finally provide an exit at NTA. VG1 is a perfect case in point, with massive underperformance, a significant Total Expense Ratio, and no feasible way to hedge out the exposure.

Still holding on for an VG1 exit at NTA, but was it worth it?

Details:

The aim of these "Cheat Sheets" is to reveal how difficult it is, even with genuine expertise, to outperform cheap index funds with CEFs in the long run. I reveal just one important but little known insight for each ASX CEF. It is not the only insight; nor always the most important one. I vary the insights to show how unnecessarily complex and hazardous CEF investing is. For detailed ASX CEF insights and expert services please use the Contact page.


1. VG1 Key Facts

30 Sep 2023:

Share Price: $1.575. Pre-tax NTA: $1.86. Post-tax NTA: $1.95 Discount: 15%

Overall, from inception at 30 Sep 2017 to 19 Oct 2023, VG1 has delivered an appalling Total Shareholder Return (TSR) of -0.34% annualised compared to 11.48% for VGS:


In six years, VG1 had one good year: FY2020-21 (when all equity returns were elevated). It made the most of it, boasting of 9.8% annualised returns to 30 Jun 2021, and claiming VG1 was now replicating its unlisted Master Fund with a supposed return of 13.8% annualised after all expenses:


Yet, now we know how it's turned out after six years: a -0.34% annualised TSR and massive underperformance of a cheap, global index fund like VGS. For over six years, the annual holding cost of VG1 has been over 11%!

In other words, if you picked a timeframe during this six year period to try and capture the discount reversion from a VG1 exit near NTA, you would have to hope to make more than 11% per year in discount capture.

The problem is VG1's max discount was 22%; the 5yr average is 11% and 3yr average is 16%. The holding cost underperformance is way too high unless your timing is perfect!

Bell Potter Weekly NTA Report - 6 Oct 2023

If you entered at the 3yr average of 16% (or the 30 Sep 2023 discount of 15%), and managed to time exit at NTA in just a year, you would likely make only a few percent return due to the massive underperformance holding costs. This pitfall is never mentioned by the CEF boosters who claim that discount reversion (especially from exits) is a key attraction of ASX LICs and LITs.

Of course, if you entered at any of the prior times in VG1's 5yr discount history, when the discount was supposedly going to be resolved in the near term, you would have significantly underperformed as the holding costs have overwhelmed any possible discount capture return.

For example, Oct 2021, when activist pressure forced VG1 to bring in MA Moelis to think "about potential structural options for the LIC" which was trading at a 15% discount.

VGI Partners calls in MA Moelis for advice

Two years on and no structural change was even seriously considered. Since Oct 2021, the captive fee revenue has been enormous while VG1 delivered a 2yr TSR of -6.48% annualised compared to 5.65% for VGS. Thus, underperformance holding costs of over 12% annualised. The cumulative underperformance holding costs are much higher over the 2yr period: over 25%!!


2. VG1 Key Insights

Fees are the barrier to CEF exits near NTA being facilitated

Firstly, let's look at the massive barrier to VGI Partners (and now Regal Funds Management since their merge) converting VG1 to an Open End Fund to enable exit at NTA. The fees to the manager have been between $8.6m and $41.7m each year:

Sep 2017 to Jun 2018: Manager Fees: $8,632,649. Expenses: $9,677,805. Net Assets: $583,706,941. Total Expense Ratio: 3.1%

Jul 2018 to Jun 2019: Manager Fees: $14,810,766. Expenses: $13,796,218. Net Assets: $941,937,304. Total Expense Ratio: 3%

Jul 2019 to Jun 2020: Manager Fees: $15,301,739. Expenses: $8,762,926. Net Assets: $892,392,253. Total Expense Ratio: 2.7%

Jul 2020 to Jun 2021: Manager Fees: $41,704,114. Expenses: $7,506,842. Net Assets: $983,926,676. Total Expense Ratio: 5%

Jul 2021 to Jun 2022: Manager Fees: $13,216,699. Expenses: $9,936,171. Net Assets: $731,115,888. Total Expense Ratio: 3.2%

Jul 2022 to Jun 2023: Manager Fees: $10,154,401. Expenses: $30,994,271. Net Assets: $702,481,512. Total Expense Ratio: At least 2% (I've offset interest income vs interest expense but the annual report is unclear about this offsetting)

In total, over $103 million in fees were paid in less than six years. If VG1 converted to Open End, there would be a flood of redemptions given performance has been so poor. Unless there was some radical shift to outperformance, I expect the fund size would shrink quickly to less than a third the current size.

The performance fee is the real killer in such funds, but due to VG1's underperformance only made a major impact in FY 2020-21. Thankfully, VG1 has a high water mark such that performance fees can't apply till the NTA gets back to where it was when performance fees were last charged.


The more "active" and "sophisticated" the fund, the higher the expenses and greater the pitfalls (you're moving away from simple asset and factor exposure)

As you can see above, other expenses have typically contributed about 1.5% to the Total Expense Ratio. The more active and sophisticated funds like VG1 have expenses like "Dividends on short positions" and "Interest expense." VG1's interest expense rocketed from $6.5m the year before to $27.3m in FY2022-23, even though Net Assets were down. This appears to be offset by Interest Income rising from $1.15m to $27.9m.

Interest rates went up significantly but that doesn't explain much. One would think careful management of such enormous changes in Income and Expenses would be discussed by VG1 in its annual report and with investors. Yet, this hasn't occurred.

As of 30 June 2023, VG1 had cash of $170m and $177m of liabilities in securities sold short. There were no other disclosed borrowings. Net equity exposure at 30 Sept 2023 was 51% and average net exposure for FY 2022-23 was 59%.

VG1 Sept 2023 Monthly Report

As with all ASX CEFs, fund managers charge fees on Net Assets (unethical managers even charge on Gross Assets), so their fees are unaffected by the discount their CEF trades at. VG1's net exposure since inception has been less than 65% so investors are paying a substantial amount of the $103 million in fees for the short selling and to hold a significant cash balance. VG1's Total Shareholder Return results indicate its short selling has been unsuccessful in contributing positively to returns.

VG1 has not provided a return analysis annually or since inception for its short selling. Yet, given how critical this is to overall performance (current short exposure is 47% and it has much greater costs than simply holding cash or being long only), it's a very poor decision to maintain this short-selling level and strategy without evidence of it being worthwhile. I should note that VG1 has indicated it has switched some short exposure from single stocks to indices in this period, but there is never any regular update as to the proportion. Shorting broad indices (e.g. NASDAQ, S&P 500) in a long-short fund is an admission the fund has no short alpha in individual stocks, which is why it is typically not disclosed transparently to investors.

One of the key pitfalls of Closed End Funds is that if they trade at significant discounts (as most ASX ones do), investors can no longer readily manage their exposure by selling and buying. VG1 investors may have initially thought its manager could more effectively time exposure levels and de-risk through expert management of single stock shorts, but the underperformance evidence indicates it can't.


VG1's active bets on currency have cost investors significantly

Unlike the vast majority of international share funds investing with US dollar exposure, VG1 actively decides to make currency bets that have great effects on returns. Typically, Australian funds don't hedge USD exposure as when equity markets sell off the Australian dollar also typically goes down against the US dollar. No investor (no matter how skilled) can time markets in the short-run, so trying to actively position currency exposure for sell-offs or bull-runs is a fool's errand.

Unfortunately, for VG1 investors, it has mostly made incorrect active bets to hedge its foreign currency exposure (mostly USD) back to Australian dollars. We can readily see this by switching the benchmark from VGS (unhedged) to VGAD (AUD hedged) for the two return periods we considered prior (inception to 19 Oct 2023; 2yrs to 19 Oct 2023):


For 30 Sept 2017 to 19 Oct 2023, the difference between VGS (11.48% annualised) and VGAD (7.46% annualised) is 4%. Thus, the cumulative difference over more than six years is over 25%!

For the last two years, the difference between VGS (5.65% annualised) and VGAD (-2.47% annualised) is over 8%. The cumulative difference of this active currency bet in just two years is over 16%!

The above analysis indicates the greatest proportion of VG1's underperformance comes from its active bets on currency. However, VG1 hasn't provided any data to justify its continuation in taking these active currency bets. VG1 holders or traders going forward may take some comfort from realising this was the major source of underperformance. The extent of this AUD/USD slide is unlikely to continue, but it's clear a lot more explanation of the significance for investors of VG1 currency positioning is required.

The myths of active management often work against objective and timely reassessments of active strategies. If it was data-driven, since Feb 2021, VG1 could have reconsidered its full hedging into AUD at multiple points. However, I suspect sustaining the myth that it can outperform via strategies like currency exposure means it is less likely to change direction or admit it has no edge and give up. Perversely, if its currency bet was just hidden in a black box (no active management reputational consequences for terminating it), it's more likely it would have been terminated. It's certainly more likely that objective data and rules may be used to limit conviction bets.

This currency exposure comparison also sheds light on what a signficant proportion of investor returns in overseas assets it composes. Investors can't rely on funds to provide sufficient data (VG1 certainly hasn't). For all investments (LICs, LITs, ETF) listed on the ASX or delivered in Australia (e.g. unlisted Australian funds), Sharesight does not provide a "Currency Gain" return as its incorporated within the fund. The only way to see currrency impacts is to compare with the few hedged products like VGAD, or with ETFs listed overseas (e.g. Vanguard ETFs listed in the U.S. like VTI and VT).


To be continued....

This investigation of VG1's underperformance holding costs will continue. Each year, VG1 provides Investor Letters, Portfolio Presentations and Investor Webcasts. It enjoys talking about its "variant perception" and focus on absolute returns over the long run. I'm waiting for the next set of updates before adding commentary.

The problem is many professional fund managers are actually intelligent, educated and skilled (though it's typically narrow). This leads to overconfidence in their ability to beat the market. They just aren't sufficiently aware of the paradox of skill, and that they aren't an exception to it.

If VG1 does continue without major change for a few more years, I will certainly update these performance comparisons and analysis, as its approach is substantially different to a cheap, global index fund like VGS. The longer the timeframe the better. I'd expect some periods of outperformance, but, in the long run, the odds of it beating simple global equity exposure costing 0.18% TER (VGS) are miniscule.


Links:

> Advisor Perspectives - Larry Swedroe: The Nonsensical Growth of Hedge Funds