Chart with yellow background

An Analysis of the Thornhill Method — Part 2: The LIC Trade-Off

Posted by on 4 June 2019 in Investing & Asset Allocation

Welcome to the next instalment of our analysis of the Thornhill Method. In the last post, we looked at dividends and showed that there is no difference between selling shares for income and living off a dividend stream. This is because any benefit of dividends is immediately arbitraged away in an efficient market.

But some readers correctly pointed out that, although we looked at high yield shares, we didn’t specifically look at listed investment companies. So in this article we’ll take a closer look at LICs.

In this article, we’ll break down how LICs were able to maintain strong dividends during the GFC. And then we’ll take a more philosophical look at “what you must believe” in order to advocate for LICs, and whether that actually makes sense.

I’m conscious that we still haven’t modelled out the performance of a pure LIC strategy (i.e. ignoring capital gains and focusing on dividend income) and a pure ETF strategy during the GFC. We’ll do that in the next post.

What is a listed investment company?

First, a little background on listed investment companies (LICs). In many ways they’re similar to managed funds, as they are actively managed portfolios that give investors exposure to a diversified basket of assets. These assets may include domestic shares, international shares, bills/bonds, and others.

The big difference between an LIC and a managed fund is the structure of each entity. On the one hand, purchasing units of a managed fund is the equivalent to giving the fund some money to invest. And this means that exiting the fund requires the fund to sell off some assets to give back to the investor.

On the other hand, the LIC funds its portfolio during an IPO. Purchasing a share of an LIC on the share market is purchasing the ownership of part of the LIC. This gives you the right to receive dividends and sell the share to somebody else later on. No funds flow in or out of the LIC portfolio when shares are bought or sold.

Traditional LICs generally invest in companies with a strong and growing dividend stream. In a general sense, this dividend stream is passed through to investors as dividends from the LICs and typically as fully franked dividends.

It’s important to remember that there is almost as much variation within LICs as there is between LICs and other investment types. So let’s narrow our focus to the specific type of LIC that Thornhill advocates. Our criteria are:

  • Active for over 50 years
  • A conservative investment philosophy that identifies value and holds for the long term
  • Management fees less than 30bps
  • No performance fees

This article will focus on LICs that meet these criteria, such as AFI, Argo (ARG) and Milton (MLT). Let’s call these ‘Thornhillian LICs’.

How did listed investment companies perform so well during the GFC?

One of the most common claims of Thornhill advocates is that listed investment companies were able to sustain dividends during the GFC. This kind of argument pops up in a bunch of ways, here’s a couple of comments from the previous post:

The preference in the bear market is perhaps based on company dividends tending to fall less than market prices.

Spending dividends is seen by many as ‘safe’, the dividends also grow over time in line with inflation or better and even in a really bad year (or years) seldom have huge immediate drops that impact.

To start with, let’s take a cursory glance to confirm that this is correct. How did our three Thornhillian LICs perform?

  • AFI did not cut their dividend at all during the GFC! Perfect start!
  • ARG cut their dividend slightly between 2008 and 2010, reaching 17% below the previous peak
  • MLT reduced their dividend quite severely in 2009 and 2010, reaching 26% below the previous peak

So the ‘average’ (and I’ll use this term loosely with a sample of three!) Thornhillian LIC did cut its dividends. But to be fair, the cut was probably less than the capital loss over as similar period. At its maximum (i.e. from pre-GFC peak to trough) the index dropped by over 50%!

We will use AFI as our test case

Of the Thornhillian LICs that we selected above, it’s clear that AFI is the best performer.  And to silence any cynics, in the rest of the article we’ll assume that AFI represents a typical Thornhillian LIC. This actually biases our analysis in favour of LICs. No straw men here!

In the chart below you can see AFI’s dividends per share between 2005 and 2018. And it’s true, they held it steady at 21c between 2007 and 2012.

But now we need to figure out how were they able to maintain their dividend payouts during the GFC. We will test two hypotheses:

  1. LICs were able to maintain sufficient earnings during the GFC, so they didn’t need to cut dividends at all.
  2. LICs didn’t maintain sufficient earnings during GFC but retained prior earnings on their balance sheet that were used to ‘beef up’ dividends during the GFC.

Hypothesis 1: LICs were able to maintain sufficient returns during the GFC

We know that LICs invest their funds in other companies with the aim of building a large and growing stream of dividends to pass onto investors. We also showed in our last post that ASX-listed company dividends were down during the GFC and stayed below the previous peak for years.

This means that LICs would have experience a reduction in earnings during the GFC (plus a fall in their trading portfolio). If this fall is significant enough, ceteris paribus, the LIC would need to cut their dividend.

We can see in the chart above that, although AFI was able to outperform the market during the GFC, they still suffered from significant losses in 2008. This was due to a reduction in dividends they received from companies they invested in.

So the next question is whether this affected their ability to sustain their own dividend in subsequent years.

The chart above compares AFI’s dividends per share and earnings per share between 2008 and 2010. As we can see, in 2009 and 2010 earnings per share was lower than dividends per share. The implication being that they were not able to generate enough earnings to cover the amount of capital returned to investors via dividends.

Hypothesis 2: LICs were able to artificially smooth returns

So we know that the best performing LIC during the GFC (AFI) was not able to generate sufficient earnings to maintain its dividend throughout the GFC, so now let’s figure out how they overcame it.

LICs have a unique legal structure which allows them to retain earnings on their balance sheet to pay out in later years (on the other hand, managed funds must pay out all their earnings immediately). So our second hypothesis is that LICs were able to artificially smooth returns by ‘saving up’ some dividends during good times and then using that surplus to maintain dividends during the GFC.

In the chart below we extended our AFI EPS/DPS chart to include three years leading up the the GFC. 

There are a couple of things to note here:

  • During the period 2005-07 you can see that earnings per share was higher than dividends per share. This implies that AFI retained some of those earnings on their balance sheet to pay out at a later date.
  • Then in 2009-10 earnings per share fell below dividends per share. This means that AFI paid out more than they earned those years. In other words, they dipped into the retained earnings that they saved up in earlier years.

It’s unfortunate that AFI only adopted the relevant AASB accounting standards in 2009, as we can actually see the growth in retained earnings after that date. Here’s a peak at what it looks like:

Implications

What does this all mean?

What do we know? We know that LICs are actively managed and can outperform the market, but still need to retain earnings to maintain their dividends during bear markets.

So any investor in LICs must believe that:

  1. Active management outperforms passive investments over the long term; and
  2. Artificial smoothing during a bear market is beneficial for an investor

Active management

As FIExporer noted in his great article on LICs, observing that some LICs may have outperformed the index over any period of time does not tell us anything about whether it was due to skill or luck. And as a follow-on, it certainly doesn’t tell us whether outperformance was identifiable in advance.

If we were to look at the evidence, we can see that over rolling 10 year period about 75% of Australian funds underperform their benchmark. The chart below is from Vanguard’s highly informative paper on indexing versus active management.

Personally, I’m not a proponent of active management, even for LICs with low fees, a long history, and a conservative approach.  There are a number of reasons for this, including:

  • I am an empricist at heart and the data tells me three things. Firstly, most funds managers will underperform the index in any given year. Secondly, fund managers that outperformed the index this year are unlikely to outperform it again next year. And thirdly, I am unreliable at picking which fund manager will outperform at any time!
  • A conservative approach is no panacea to active management. There’s no way to eliminate manager risk (i.e. the risk that a manager selects the wrong stock or doesn’t select the right stock). 
  • A long history is indicative of very little, as by definition all LICs with a long history will have be well-performing due to the effect of extinction. And, of course, most managed funds do perform well… until they don’t!

Artificial smoothing

I’ve also heard that argument that artificial smoothing is beneficial, as it guarantees that we won’t waste or misuse our income when times are good, leaving us with nothing for when times are bad. 

Some of may be aware of the time value of money. The time value of money states that money today is worth more than money tomorrow. And importantly for the current discussion: money tomorrow is worth less than money today.

There are two reasons for this:

  1. There is a true opportunity cost of not having the money to invest. It’s like leaving your investment money in a transaction account that generates 0% interest when it could be in bonds or equities earning 4%+
  2. Inflation will slowly but surely erode the purchasing power of money not being invested. So you’re not just making 0%, you’re making negative 1-3% (probably)!

So to bring it back to our discussion, instead of the LIC paying out dividends, it may decide to wait for the next bear market. This means that we miss out the opportunity to invest that money and the potential years of compound interest associated with that investment. The so-called benefit of dividend smoothing is really a cost.

Now what?

It’s clear that, although some LICs performed well during the GFC, that performance is contingent on two principles that aren’t entirely rational. 

The outperformance of LICs during bear markets is dependent on them outperforming the market in earlier periods and retaining those earnings. Without that outperformance due to active management, retaining earnings will result in the LIC underperforming the index and the investor losing out in net present value terms.

The information in this website and the links provided are for general information only and should not be taken as constituting professional advice. Ordinary Dollar is not a financial adviser. You should consider seeking independent legal, financial, taxation or other advice to check how the website information relates to your unique circumstances. Ordinary Dollar does not guarantee the accuracy of any information on this website, and is not liable for any loss caused, whether due to negligence or otherwise arising from the use of, or reliance on, the information provided directly or indirectly, by use of this website.

There are 14 comments on this article

  1. Avatar for Dan Montgomery

    Dunno

    Hi Dan

    It is only the retention of the franking credit itself that has any time value. The cash not paid out remains invested by the LIC.

    The Math that supports passive investing beating active investing on average, is due to expenses of active investing. With the old school LIC’s having management costs and turnovers not far different to capital weighted ETF’s they should have pretty much the same probability of outperforming as underperforming. But no doubt LIC’s ara a active allocation and should be taken with eyes wide open to the possibility of differing returns to the market.

    Before ETF’s, these grandfather LIC’s were the best option for index like diversification. Their payout policy of smooth dividends helped people gauge their own personal spending capacity. In other words, they were and continue to be a substitute for total return perspective which leads towards capital weighted index diversification and safe withdrawal rates to determine spending capacity.

    How is your SWR calculator project going? A simple and robust understanding of SWR and how to calculate it is needed before total return investing makes sense to lots of people who don’t particularly find the nuts and bolts of financial efficiency interesting.

    • Avatar for Dan Montgomery

      Dan Montgomery

      Thanks for the feedback Dunno! When I took a look at the AFI annual reports it didn’t look like they were investing retained earnings. Where did you see that? If I can find some evidence then I’ll need to update the article!

      The SWR is broadly finished. It just needs some instructions and some UI improvements. I’ll look to launch it right after this LICs series 🙂

      • Avatar for Dan Montgomery

        Dunno

        Compare cash account to retained earnings in the financial statements.
        They can also pay dividends from the realised capital gains reserve. These historical realised gains not yet paid must also be re-invested given the size of cash account compared to the reserve.
        They can raise the cash for future dividend smoothing in the event of income shortfall by either sales of investments, issuing new shares via DRP, SPP’s etc or even from borrowings if that is part of their investment strategy.

        The franking credit balance is the practical limit to dividend smoothing as paying an unfranked dividend is functionally the same as a return of capital and no real benefit to shareholders as they could just sell themselves if they wish to consume capital. Its my observation that most LIC’s limit smoothing to franking ability, not retained earnings and realise gain reserve exhaustion.

        Retention of profits is a “potential” strength of LIC’s so long as they re-invest well, just like any other company.

        Looking forward to seeing your SWR calculator.

  2. Avatar for Dan Montgomery

    Russell

    Loving the analysis Dan, keep up the good work!

    On my FIRE journey, which is still very new, I got temporarily sucked into the LIC approach and invested some money. I will probably stick with index funds from here on out for the very reasons you’ve mentioned.

    Do you think there is a danger, given the shift in mentality towards index investing, that the returns of index investing will reduce over time as more and more institutions and individuals invest in the index as a whole as opposed to trying to “pick the winners”? I admit my understand of how this all works is hazy. Interested to hear your opinion! 🙂

  3. Avatar for Dan Montgomery

    Dave @ Strong Money Australia

    Agree with the comments above. On your two points…

    Dividend Smoothing.
    As above, retained earnings are reinvested, only franking builds up on the company account, just like any company that pays out less than 100% of earnings. This is not reliant on outperformance, it is based on the cashflow and payout ratio of the LIC.

    Active Management: No question there is risk of underperformance. Vanguard also have a paper showing by far the best predictor of active fund performance is cost, and that low cost active funds generally fare much better than the rest. This is not at all to say they’ll beat the market, just that it’s probably not as bad as some make it out to be.

    Regardless, you’d probably find that the goal of most investors in these funds is not to beat the market. Most want to be invested in a low cost equity fund with a relatively reliable income stream to live on, either now or in the future. And if the desire for a simple approach and more predictable cashflow means they end up lagging the market a bit, it’s probably not the end of the world.

    The average of the dividend cuts in the example was 14.3%, so a decent bit less than the price falls of 55%. That can make a big difference to the psyche of an investor who is living off income during a GFC and help them to stay the course. These vehicles are not for everyone, maybe even not for most people. But this dividend approach and low cost LICs work for some, so that’s probably what matters.

    • Avatar for Dan Montgomery

      Dave @ Strong Money Australia

      ‘Agree with the comments above’… obviously I meant the first comment by ‘Dunno’ 😉

    • Avatar for Dan Montgomery

      Dan Montgomery

      Thanks again for your comments Dave, I do like hearing a different perspective. As I mentioned to Dunno, do you know where you saw that LICs invest retained earnings? It didn’t seem that way from reading the financial statements but I’m no LIC auditor!

      I agree with you that there’s a psychological benefit to a stable dividend stream. I just think of things in terms of financial risk and reward, and LICs don’t quite hold up for me.

      I have a question about LICs and the dividend approach. How do you know when your dividends are sufficient to retire? No doubt they’ll be cut during a crash, so how much buffer is recommended by Thornhill?

      • Avatar for Dan Montgomery

        Dave @ Strong Money Australia

        I believe Dunno has explained better than I could! Cash balance is typically 2-5% of portfolio over the years, everything else is invested/reinvested. Otherwise they’d end up with a growing cash balance each year until a downturn, that’s not what happens.

        As a simplified example, if their portfolio yields 4% and dividends are cut by 25% that year, they’ll need to use 1% from the portfolio’s cash balance and tack on equivalent franking from the franking account balance, if they want to top up the dividend and avoid cutting. I’m no accounting wiz either, but that’s my understanding.

        When your dividends are a bit more than your spending, that’s probably fine. Some people will like to have a nice gap there before retiring, and I’d probably be cautious if we’d just had a run of large dividend growth (double digit), but generally when income exceeds expenses is the idea.

        Personal cash buffer is different for everyone, kind of like a bond allocation perhaps. Very much dependant on the individual’s ability and willingness to spend less, earn additional money etc. From memory Thornhill suggests something like 2-3 years income as a cash buffer, combined with spending restraint should be enough to plug the gap. Some may want more, others less.

  4. Avatar for Dan Montgomery

    The FI Explorer

    That’s a really good article, and I can’t wait for the next instalment. I really admire the empirical effort and robustness you put into these posts, which distinguishes them from some the debate in this area.

    On Dave’s points above, I think a lot hangs on what the harmless sounding phrase ‘lagging the market a bit’ translates to, and for whom.

    For an income seeking retiree with a large portfolio and a margin of safety, that is probably fine.

    For a younger FI seeker with a time frame of 10-50 years or more on their portfolio, even a small ‘lag’ is a big impediment to compounding occurring and potentially even reaching FI.

    And as the academic evidence mentioned in my analysis show, active management typically leads to a large lag, not a small lag, in the order of 2.5-3.0% pa. The snapshot of some of the Thornhill LICs performance included also reinforced that finding.

    A performance deficit of those kinds over a long time is not a small issue, it’s a real problem. This is why I would question LICs benefits for those still in the accumulation stages of FI, and some suggestions along those lines from the Barefoot Investor.

    • Avatar for Dan Montgomery

      Dan Montgomery

      Thanks mate, appreciate the post. I agree with small differences compounding into big differences decades down the track.

      I’ll try to model out a Thornhill vs index strategy during the GFC and see whether they differ. We will probably have to make some assumptions but it will be an interesting exercise!

      • Avatar for Dan Montgomery

        Steve Dodds

        Numbers can be deceptive depending on timeframe. And, alas, I can’t seem to post a graph here. But according to Sharesight an investment in AFI made on April 1 2007 had a capital gain of .56% and dividends of 4.45% for a total of 4.8%.

        Same investment in STW had -.35% CG , 3.7% dividends and TR of 3.43.

        Change the start date to 10 years or 5 years or less and the index wins.

        However, AFI is considerably less volatile.

        This may be chance. But according to the Sharesight charts AFI (or your LIC proxy) went down less during the GFC and went up less during the bull market that followed.

        The real share price underperformance has been in the last couple of years.

        A Thornhillist would say that is exactly what is supposed to happen.

        A TRist would say you’ve missed out on some returns.

  5. Avatar for Dan Montgomery

    Barry

    Great article, enjoyed the read and look forward to the next installment

    The below was also good, its hard to compare like for like

    https://cuffelinks.com.au/lic-performance-reporting-inadequate/

    Lastly, I was curious what is an empricist :o)

  6. Avatar for Dan Montgomery

    Aussie HIFIRE

    Another great post Dan. I must confess that I wasn’t aware that some of the bigger LICs had cut their dividend in the GFC as AFI tends to be the one that gets mentioned the most. In which case if you’re not getting the benefit of dividend smoothing, and you’re likely to underperform the index on a long term basis, plus the index is paying out roughly the same dividend and franking credits as the LICs, why invest via LICs?

  7. Avatar for Dan Montgomery

    number 5

    Great read Dan.

    I have ARG, MLT and VAS in my portfolio. There is a feel good factor when buying a LIC at a discount. BUT over time i do see more and more of my new money going into VAS. I’m still not completely convinced one vehicle is better than the other, so ill keep holding both!

Responses to Aussie HIFIRE

Click here to cancel reply.