Category: MAPF

The market experienced a wild ride in July, with the PerpetualDiscount index down 8.85% for the month at nadir on July 16, subsequently recovering to a loss of “only” 3.34% as the bargain-hunters swooped in on some very juicy yields. See Best & Worst Performers, which shows results for both parts of the month as well as the whole, and Index Performance, July 2008 for comparison with other subindices and the two exchange-traded funds.

The fund, with its heavy weighting in PerpetualDiscounts (see MAPF Portfolio Composition, July 2008), was not immune to the carnage and declined 2.31% on the month, before fees but after expenses.

Returns assume reinvestment of dividends, and are shown after expenses but before fees. Past performance is not a guarantee of future performance. You can lose money investing in Malachite Aggressive Preferred Fund or any other fund. For more information, see the fund’s main page.

The yields available on high quality preferred shares remain elevated, which is reflected in the current estimate of sustainable income.

It is very gratifying to see the sharp increase in expected income per unit – if there was no trading in the fund, this would be a constant number. While it is disappointing to see the net asset value per unit decline, it is the expected income per unit that will determine the long term performance of the fund as the market stabilizes.

I must point out, however, that the expected income is a little skewed this month due to the extraordinarily high yield available on WFS.PR.A, which was purchased in quantity during July. At month-end, this position was valued to have a yield of 9.50% (as a dividend!), but the security matures in three years time – so, while the calculation shown is accurate as far as it goes, as a long-term indicator it is expected to decline in three years when WFS.PR.A is redeemed and the proceeds reinvested in a security that will not necessarily yield 9.50%.

If WFS.PR.A had been sold at market value at the end of July and the proceeds reinvested proportionately in the existing portfolio, the “Securities Average YTW” in the table above would have fallen about thirty basis points (0.30%) to approximately 6.66%; this would result in an estimate of “Sustainable Income” of $0.5428; less than is reported, but still a substantial increase from previous figures. It is hoped, of course, that the market will shortly recognize the merits of WFS.PR.A and bid up the price until the yield is – according to me! – more reasonable, which should allow the fund to take a good-sized capital gain when swapping it for another issue with upside potential.

So, despite the poor price performance in July, we must remember that we are fixed-income investors. The expected annual income per unit (these are shown gross of fees and expenses) continues to show an upward path … and it is the income that makes the asset class worthwhile.

I should emphasize, however, that the fund does not explicitly seek to maximize this number. Yield on the portfolio will be given up when it is possible to exchange it for something else that is attractive: credit quality, say, or retractibility. Over the very long term, however, it is the prime objective of fixed income management to maximize the income received from a given amount of capital.

The fund’s heavy weighting in the underperforming PerpetualDiscounts hurt performance for the month; within that group, a relatively large exposure to CM issues was another negative factor. The fund was able to mitigate the effects of these two allocations by frequent trading within these two groups; trading was very heavy during the month, representing approximately 100% of portfolio value.

As may be seen in the table above, there was considerable chaos in the downward movement of the CM issues, which was mitigated by opportunistic trading between the issues. These trades did not, in and of themselves, change the portfolio’s credit risk or have a material effect on any element of the portfolio’s overall risk profile. Opportunistic trading is what MAPF is all about!

Overall, the market has normalized somewhat since the end of June. The extremely strange – and theoretically unsupportable – relationship that existed a month ago between prices and yields has basically reversed itself – although there remain small pockets of strangeness and always will! Theory will always reassert itself in the long run; it is simply unfortunated that in June I was caught in the underperforming end of the strangeness and hence could not execute trades to exploit the disparities.

And for the future? In the first few days of August, the market has moved higher, with PerpetualDiscounts up 0.72% in the first few days. The average yield on PerpetualDiscounts at month-end was 6.32% (as dividends), equivalent to interest income of 8.85% at the standard equivalency factor of 1.4x. This in turn represents a spread of 270bp over long corporates – and, as has been previously mentioned, the previous 10-year high was 250bp – so, some might say, we are still well within the area of a 10-year buying opportunity!

While it will take the credit crunch a great deal of time to unwind – and while I continue to disdain market timing, these elevated yields should gradually attract bond investors, which will help prices recover. But … keep your eyes on the income! I continue to apply and refine my processes for finding opportunities, such as the CM swaps shown above, to execute swaps with a favourable risk/return profile.

My next article for Canadian Moneysaver will discuss the events of the last two months in more detail and, once the black-out period has expired, will be republished here.

There was a substantial amount of trading in July, as the collapse in prices of PerpetualDiscounts in a confused market brought many opportunities to the Fund. Turnover was close to 100% for the month, but a high proportion of these trades were intra-issuer (trades between the CM issues were particularly frequent) and most others were intra-sector (PerpetualDiscounts fell at different rates).

Trades were, as ever, triggered by a desire to exploit transient mispricing in the preferred share market (which may the thought of as “selling liquidity”), rather than any particular view being taken on market direction, sectoral performance or credit anticipation.

The “total” reflects the un-leveraged total portfolio (i.e., cash is included in the portfolio calculations and is deemed to have a duration and yield of 0.00.). MAPF will often have relatively large cash balances, both credit and debit, to facilitate trading. Figures presented in the table have been rounded to the indicated precision.

The increase in SplitShares is due to purchases of WFS.PR.A, discussed below.

Credit distribution is:

The fund does not set any targets for overall credit quality; trades are executed one by one. Variances in overall credit will be constant as opportunistic trades are executed.

The increase in the proportion of Pfd-2(low) issues held is due to purchases of WFS.PR.A:

As of month-end, the move into WFS.PR.A has not borne fruit – but given the substantial yield pick-up (over 3 points!) I consider it to be only a matter of time before the WFS.PR.A experiences a substantial price increase. As of July 24, it had asset coverage of 1.6+:1, according to Mulvihill; comparable issues (insofar as anything is comparable; as the name implies, World Financial Split has a portfolio with considerably more geographic diversification than most split share corporations) are trading to yield 5.5%-5.8%. If WFS.PR.A were to trade to yield a more reasonable (to me!) 6%, the price would be about $9.85, an increase of almost 10% from the the July month-end value. Frankly, it seems like a pretty good bet, particularly given the decrease in interest-rate risk when trading from perpetuals to a split-share maturing in just under three years!

Liquidity Distribution is:

MAPF is, of course, Malachite Aggressive Preferred Fund, a “unit trust” managed by Hymas Investment Management Inc. Further information and links to performance, audited financials and subscription information are available the fund’s web page. A “unit trust” is like a regular mutual fund, but is sold by offering memorandum rather than prospectus. This is cheaper, but means subscription is restricted to “accredited investors” (as defined by the Ontario Securities Commission) and those who subscribe for $150,000+. Fund past performances are not a guarantee of future performance. You can lose money investing in MAPF or any other fund.

A similar portfolio composition analysis has been performed on CPD as of May month end; it should be noted that the underlying TXPR index has been rebalanced and I have not yet fully analyzed the changes. While the changes affect the allocation to the different sectors, I do not believe the credit or liquidity metrics will have changed much.

• MAPF credit quality is superior
• MAPF liquidity is comparable
• MAPF Yield is higher
• But … MAPF is more exposed to PerpetualDiscounts
• MAPF is less exposed to Fixed-Resets

The fund experienced disappointing returns in June, as the market didn’t just collapse, it collapsed without retaining a normal degree of internal consistency.

Returns assume reinvestment of dividends, and are shown after expenses but before fees. Past performance is not a guarantee of future performance. You can lose money investing in Malachite Aggressive Preferred Fund or any other fund. For more information, see the fund’s main page.

The yields available on high quality preferred shares remain elevated, which is reflected in the current estimate of sustainable income.

So, despite the poor price performance in June, we must remember that we are fixed-income investors. The expected annual income per unit (these are shown gross of fees and expenses) continues to show an upward path … and it is the income that makes the asset class worthwhile.

I should emphasize, however, that the fund does not explicitly seek to maximize this number. Yield on the portfolio will be given up when it is possible to exchange it for something else that is attractive: credit quality, say, or retractibility. Over the very long term, however, it is the prime objective of fixed income management to maximize the income received from a given amount of capital.

When we look at the MAPF Portfolio Composition for May, 2008, we see that the fund was invested almost entirely in PerpetualDiscounts; this overall analysis is unchanged in the portfolio composition for June. Given the sharp decline in the market in June, it should therefore come as no surprise that the the fund underperformed, but even so … PerpetualDiscounts declined 5.31% in June, so there is a further source.

I have made some data available with this post:

• PerpetualDiscount Index, June 30
• PerpetualDiscount Index, May 30, with June Performance
• Excel Spreadsheet, used to prepare the charts and regressions in this post

When we look at the June performance of the PerpetualDiscount Index as of May 30, the first thing we notice is that the non-financials did relatively well:

BAM is perceived as a financial, but is classified as an “Industrial” by the TSX. Whichever way you slice it, financials did extremely poorly relative to non-financials, and this factor alone is enough to explain the underperformance of MAPF relative to the PerpetualDiscount index – the fund has a position in BAM.PR.N, but the other elements of its PerpetualDiscount exposure is unequivocally financial.

The remainder of the analysis will examine only the financials.

I have previously noted that the more deeply discounted perpetuals underperformed in the period 5/30 to 6/13. The following chart shows that this effect persisted through monthend:

A regression of the May 30 Price against June return for the issues rated Pfd-1 shows that the effect is significant: 28% of the variation is explained by the equation:

Return = -19.77% + 0.65*Price %

Thus, the expected return for a Pfd-1 financial issue in June with a May 30 price of $20 would be -6.77%, while an initial price of $25 would predict a June return of -3.52%. Surprisingly, this effect only hits the Pfd-1 issues … a similar regression for the Pfd-1(low) issues produced no result, while there are not enough data to examine in the Pfd-2(high) and Pfd-2(low) grades.

The price/return relationship for Pfd-1 issues is not just unexpected, it is ludicrous – as I have been harping on since mid-month. I will stress again that my expectations are not based on mere observations of historical relationships – these expectations are at the heart of fixed-income theory.

When selecting a fixed income portfolio, one always bears three scenarios (at least!) in mind – rates generally rise, rates generally fall, rates are essentially flat. More sophisticated analysis is, in many ways, simply an elaboration of these basic assumptions about the future. So, when we compare a low-coupon, low-price PerpetualDiscount to a high-coupon high-price PerpetualDiscount, we come up with the following implications of each scenario (for more information, see my article Perpetual Hockey Sticks):

• If rates go up, the price of each instrument will decline by approximately the same percentage.
• If rates go down, we expect the low-coupon, low-price issue to increase in price more, since it has more room to increase before the holder has to worry about his capital gains being called away.
• If rates remain the same, the higher yielding issue will have a better return. Since the two issues are equal in first scenario, and the low-coupon issue wins in the second, we expect the high-coupon, high-price issue to have the better return in this scenario

Or, to put it another way, investors should demand a higher yield on the high-coupon issues, in order to compensate for the possibility that they might lose out on capital gains if rates decline and the issuer calls the issue for redemption.

As is always the case in preferred share analysis, there is never enough data to show many perfect illustrations of various points – there are just too many cross currents, in terms of credit quality, issuer, coupon, redemption terms … the list is endless. However, there are two issuers of Pfd-1 quality with PerpetualDiscount issues outstanding that cover a broad enough range of coupons to be interesting: CM and RY. The following charts show the May 30 and June 30 yields, plotted against their coupon. It should be noted that if there was no convexity effect (which I would consider the limiting case) the plots should always be flat; a normal convexity effect should show that yield increases as the coupon gets larger, as compensation for the decreased room for capital gains before the holder has to worry about a call.

Clearly, the curve change from “normal” at the end of May to “abnormal” at the end of June. Please note that I have deliberately not referred to the June relationship as “inverted” … while normal yield curves can invert (when short rates are higher than long rates), this is an economically reasonable relationship under the correct economic conditions. There are no conditions in which high-coupon perpetual discounts should yield less than low-coupon perpetual discounts, in the absence of special, issue-specific factors, as I noted on June 27:

• A big difference in term to call
• A big difference in liquidity
• A big difference in other terms of the issue (e.g., voting rights, restrictive covenants, etc.)

None of these features is applicable to the CM and RY issues.

Well … despite the fact that the changes in relationships are impossible, they happened anyway. Markets have a way of doing that, just to remind us that we don’t know everything! So why did all this happen?

I suspect that:

• The overall decline in PerpetualDiscounts is due – at least in part – to the Bank of Canada June 10 decision to keep the overnight rate constant. There are many who imagine that there is only one interest rate … if the Government of Canada overnight rate is flat-to-increasing, they think, the same must apply to long term corporate rates. Therefore, they sell, in expectation of future price declines, which become (for a while, anyway, in a small enough market) a self-fulfilling prophecy
• The huge difference in returns between financial and non-financial perpetualDiscounts implies that Fear of Banks is a major factor. It is my feeling that such fears are misplaced. While Canadian banks are certainly not unscathed by the credit crunch, they’re not exposed to the full force of leveraged positions in sub-prime paper either! I do not feel that the actual chance of default by any of Canada’s banks has increased in any kind of material way.
• The abnormal coupon-yield relationship points to retail. Selling something because the price is down, rather than on an objective evaluation of risk/return, is never a winning strategy, but one can always count on retail to do the wrong thing.
• The fact that this effect is most pronounced in the Pfd-1 issues (and barely visible in Pfd-1(low)) points to not just retail, but small-time retail at that. If one can only hold one or two preferreds (due to constraints of portfolio size), it only makes sense to hold tip-top quality – I’ve made that recommendation myself. If these small holders are eager to dump (a portion of) their preferreds, they don’t have a lot of choice as to which ones

What does it all mean? June was a bad month. The fund is overweighted in high quality, low-coupon, financial PerpetualDiscounts and this was precisely the wrong spot to be. HIMIPref™ is a statistically based system and will not work well every time. However, trading continues, based on expected incremental returns and I have every expectation that good results from this trading will become visible in terms of fund return as the market normalizes.

Trading started off slowly in June, but two bursts of activity from June 10-13 (when spreads started widening) and June 27-30 brought portfolio turnover up to about 65%.

Trades were, as ever, triggered by a desire to exploit transient mispricing in the preferred share market (which may the thought of as “selling liquidity”), rather than any particular view being taken on market direction, sectoral performance or credit anticipation.

The “total” reflects the un-leveraged total portfolio (i.e., cash is included in the portfolio calculations and is deemed to have a duration and yield of 0.00.). MAPF will often have relatively large cash balances, both credit and debit, to facilitate trading. Figures presented in the table have been rounded to the indicated precision.

Credit distribution is:

The fund does not set any targets for overall credit quality; trades are executed one by one. Variances in overall credit will be constant as opportunistic trades are executed.

The slight decline in credit quality is the result of a move from banks to insurers; in overall terms that do not reflect specific trades, a May month-end position of 7% BMO and 32% RY issues became 8% POW (Pfd-2(high)), 11% GWO (Pfd-1(low)) and 18% PWF (Pfd-1(low)).

The first trade I will examine comprises the largest single piece of the series of trades – unfortunately, it would have been better to delay:

Unfortunately, sometimes you’re just going to be too soon – particularly when you are trading on random noise … sometimes the random noise gets louder than usual, sometimes (in the worst case scenario) the so-called noise turns out to be a trend. HIMIPref™ simply plays the odds: with enough trades, actual results will reflect the statistics. At time of writing, RY.PR.C is bid at $19.05, while GWO.PR.G is bid at $20.84, so the trade results have improved since month-end – although it’s still underwater.

Liquidity Distribution is:

MAPF is, of course, Malachite Aggressive Preferred Fund, a “unit trust” managed by Hymas Investment Management Inc. Further information and links to performance, audited financials and subscription information are available the fund’s web page. A “unit trust” is like a regular mutual fund, but is sold by offering memorandum rather than prospectus. This is cheaper, but means subscription is restricted to “accredited investors” (as defined by the Ontario Securities Commission) and those who subscribe for $150,000+. Fund past performances are not a guarantee of future performance. You can lose money investing in MAPF or any other fund.

A similar portfolio composition analysis has been performed on CPD as of May month end. It is interesting to note:

• MAPF credit quality is superior
• MAPF liquidity is superior
• MAPF Yield is higher
• But … MAPF is more exposed to PerpetualDiscounts

This last factor hurt performance in June and will be discussed when the performance is posted.

Trading continued to be slow in May, with only about 30% of portfolio value being traded – and at that, much of the activity was intra-issuer. On a net basis, there was a slight shift from BMO to RY. Trades were, as ever, triggered by a desire to exploit transient mispricing in the preferred share market (which may the thought of as “selling liquidity”), rather than any particular view being taken on market direction, sectoral performance or credit anticipation.

The “total” reflects the un-leveraged total portfolio (i.e., cash is included in the portfolio calculations and is deemed to have a duration and yield of 0.00.). MAPF will often have relatively large cash balances, both credit and debit, to facilitate trading. Figures presented in the table have been rounded to the indicated precision.

Credit distribution is:

The fund does not set any targets for overall credit quality; trades are executed one by one. Variances in overall credit will be constant as opportunistic trades are executed.

Liquidity Distribution is:

MAPF is, of course, Malachite Aggressive Preferred Fund, a “unit trust” managed by Hymas Investment Management Inc. Further information and links to performance, audited financials and subscription information are available the fund’s web page. A “unit trust” is like a regular mutual fund, but is sold by offering memorandum rather than prospectus. This is cheaper, but means subscription is restricted to “accredited investors” (as defined by the Ontario Securities Commission) and those who subscribe for $150,000+. Fund past performances are not a guarantee of future performance. You can lose money investing in MAPF or any other fund.

Update: A similar analysis has been performed on CPD as of month end. It is interesting to note:

• MAPF credit quality is superior
• MAPF liquidity is superior
• MAPF Yield is higher
• But … MAPF is more exposed to PerpetualDiscounts

Update, 2008-6-3: May Performance for MAPF has been posted.

The fund had a very good return in May; very slightly behind CPD (which returned +1.42% on the month). The BMO-CM “50” index return is not yet available; this post will be updated when it is published.

Returns assume reinvestment of dividends, and are shown after expenses but before fees. Past performance is not a guarantee of future performance. You can lose money investing in Malachite Aggressive Preferred Fund or any other fund. For more information, see the fund’s main page.

The yields available on high quality preferred shares remain elevated, which is reflected in the current estimate of sustainable income.

Update, 2008-6-3 : Index performance for May has been posted. MAPF Portfolio Composition, as has a similar analysis for CPD, the exchange traded fund.

The fund had a good month in April, with substantial outperformance against its benchmark.

Returns assume reinvestment of dividends, and are shown after expenses but before fees. Past performance is not a guarantee of future performance. You can lose money investing in Malachite Aggressive Preferred Fund or any other fund. For more information, see the fund’s main page.

The yields available on high quality preferred shares remain elevated, which is reflected in the current estimate of sustainable income. The figures for YTW and Leverage Divisor were disclosed in the discussion of the April month-end portfolio composition.

It is noteworthy – well, I consider it noteworthy – that the “Securities YTW” remained constant during the month despite the increase in portfolio credit quality during the period. The intent (insofar as a purely quantitative system can be said to have an intent) is to sell that quality back to the market as soon as it is not quite so cheap.

The slight increase in “Sustainable Income” is largely due to the accumulation of one month’s worth of dividends, which have been received by the fund but not yet paid to unitholders.

I will post more discussion later.

Trading slowed down somewhat in April to about 35% of portfolio value. “Slow” is a relative term! Trades were, as ever, triggered by a desire to exploit transient mispricing in the preferred share market (which may the thought of as “selling liquidity”), rather than any particular view being taken on market direction, sectoral performance or credit anticipation.

The “total” reflects the un-leveraged total portfolio (i.e., cash is included in the portfolio calculations and is deemed to have a duration and yield of 0.00.). MAPF will often have relatively large cash balances, both credit and debit, to facilitate trading. Figures presented in the table have been rounded to the indicated precision.

Credit distribution is:

The fund does not set any targets for overall credit quality; trades are executed one by one. Variances in overall credit will be constant as opportunistic trades are executed.

Liquidity Distribution is:

MAPF is, of course, Malachite Aggressive Preferred Fund, a “unit trust” managed by Hymas Investment Management Inc. Further information and links to performance, audited financials and subscription information are available the fund’s web page. A “unit trust” is like a regular mutual fund, but is sold by offering memorandum rather than prospectus. This is cheaper, but means subscription is restricted to “accredited investors” (as defined by the Ontario Securities Commission) and those who subscribe for $150,000+. Fund past performances are not a guarantee of future performance. You can lose money investing in MAPF or any other fund.

And, here’s an example of a trade sequence that worked marvellously. As always, note that the data in the table has been aggregated and approximated; the prices shown are a fair approximation of the detailed figures, but do not necessarily represent actual execution prices. Trade detail is made public periodically via the fund web page, but details for 2008 trades have not yet been released.

April performance will be published on the weekend.

The fund’s amazing run of three superb months in a row came to a halt in March as the market swooned with the fund having overweighted PerpetualDiscount issues, the hardest hit sector. A fair bit of trading mitigated, but could not eliminate, the damage.

This was only one month, however, and the fund takes a long-term approach to the markets – it is recognized that not every month will deliver excess returns, or even every quarter. Trades are executed when there is a good probability of relative profit and in the past this has brought excess returns over time, albeit with considerable lumpiness in the timing of these excess returns.

Returns assume reinvestment of dividends, and are shown after expenses but before fees. Past performance is not a guarantee of future performance. You can lose money investing in Malachite Aggressive Preferred Fund or any other fund. For more information, see the fund’s main page.

The competition was outpaced for the quarter: the fund outperformed the closed-end fund (DPS.UN), which returned an estimated -2.26% on the month and an estimated -0.79% on the quarter, as well as the exchange-traded fund (CPD) which returned -2.90% and -1.23% on the month and quarter. Calculation details for these two performances have been posted separately.

The yields available on high quality preferred shares remain elevated, which is reflected in the current estimate of sustainable income.

It should be noted that I do not have this calculation audited in any way, so unitholders will not be able to see an explicit confirmation of these figures, although you will be able to derive the year end figure for yourselves – I will be happy to provide supporting documents for the calculation to unitholders on request. Readers should also note that the fund is indifferent to whether investment returns are in the form of capital gains or dividends – portfolio management seeks to maximize total return after tax for a notional high-marginal-rate investor based in Ontario. It should also be noted that this sustainable income figure is not targetted in any manner; it may well go down if, for instance, it is decided that quality is cheap and trades are executed to increase credit quality at the expense of yield.

For all that, though, there is a point to the calculation – it shows that in the recent past, and subject to the usual warning that historical performance is not necessarily indicative of future returns:

• Income expectations are a lot more stable than market prices, and
• the overall trend is upwards

In the year ended March, 2008, however, the total dividend distribution of $0.525925 was very close to the theoretical figure, albeit with considerable quarterly variance.

The market finished March just a little bit above the trough of November, 2007, as several new issues (NA 6% Perps, BMO 5.80% Perps, BNS Reset Perps and TD 5.60% Perps) knocked the market down considerably. I suspect that there will be something of a pause in issuance for the nonce, as the market recovers … but I’ve been wrong on these macro-calls before and I’ll be wrong again in the future! The issuers will do what’s good for their business, without worrying too much about what’s good for the marketplace.

Long term investors will be most interest in the dividend-friendly Ontario budget, which went a long way towards countering the future effects of the dividend hostile Federal budget.

The fund did considerable trading during the month, but most of this trading was simply opportunistic switching between issues with similar characteristics.