Description

Dipula is a South African Reit. It has an interesting capital structure with an A and B share. The dividend yield on the A-share is 15%, the B-share 26% and combined 18%. Investors can create a synthetic B-share which yields 29%. 

Management owns 10% of the B-shares and are motivated to unlock value. 

Business Description

Dipula owns 189 properties (GLA = 926 000m2) valued at R8.7bil. The properties are funded with debt of R3,2bil. The cost of debt is 7.8% and 60% of the debt is fixed. 

The properties generate an annual rent of R1.3bil split: Retail (65%), Industrial (15%) and Office(19%).

The A+B structure

Many South African REITS issue both A and B shares. The A shares take a first share in the distributable income and the B share gets whatever is left over.

The amount that the A-share is entitled to is defined when the share is listed. This initial dividend escalates by the minimum of 5% or CPI.

To illustrate the structure, Dipula generated distributable income of 104cps for the six months ended Feb2021. The payout ratio was set at 100% which meant that the dividend was also 104cps. The A-share was entitled to the first 59c and the B-share was left with the remaining 45c. Hypothetically, if the payout ratio was only 75%, the dividend would only be 78c. The A-share still gets 59c and the B-share only gets 19c.

While a decline in the dividend hurts the B, the reverse holds when earning are increasing. The REIT sector in South Africa had been in a bull market for over 15 years and nobody could imagine declining earnings when these structures were put in place.

Dividend Yield

Annualizing the 59c and 45c implies a 118c and 90c dividend for the year. As the LDR for the 1H2021 dividend is next week (June,2 2021) the clean price is used to calculate the dividend yield.

Obviously these are very attractive dividend yields, but, are meaningless if they can not be sustained.

Sustainability of the Dividend

Covid has put the earnings power of Reits to the ultimate test and as the table of six monthly results shows Dipula passed this test with flying colors. 

(For perspective a national lockdown lasted from March 27, 2020 to May 1, 2020. From then restrictions were gradually relaxed until a second wave hit in December. On December 29, 2020 restrictions were tightened until March 1, 2021.)

Observations

The  six months ended Aug2020 coincided with the most severe Covid restrictions. Rent collected in this depressed period fell from R661mil to R587mil mostly as a result of discounts of R43mil. In the following  six months, rents bounced back to R668mil (net of R8mil discounts) slightly above pre-covid levels. 

Operating profit was flat, but, distributable income grew by 11.8% from a pre-covid R254mil to R284mil.

The resilience of the income during these restrictive times is impressive and is largely due to the interplay of the following dynamics.

• Traditionally South African leases are negotiated with annual escalation clauses. Dipula’s current portfolio of leases has an average escalation rate of 7%. This helped mitigate the loss in rent due to increased vacancy (7.6% vs 5.8%) and negative rental reversions.

• Renewed Leases (44 000m2): Rent reversions on renewed leases for the six months ended Feb2021 were -4.8%. The reversions indicate some stress, but this will be temporary as the escalation rate of these renewals was 7%.  

• New Leases (25 000m2): The initial rent was in line with the old rent with an average escalation of 7.7%. 

• Only 60% of the debt (R3.2bil) is fixed. South African interest rates have been declining driving down Dipula’s cost of debt from 9.3% to  7.8%. The interest bill should continue to decline as Dipula recently renewed a debt facility of R590mil at 6.5%.

• Covenants are not in play, ICR = 3.2x (required 2x), LTV = 35% (required 45%)

The above factors are very encouraging and it seems likely that Dipula will generate sufficient income to maintain distributable income at the current level.

Payout Ratio

What is less certain is the payout ratio. Despite the fact that Dipula generated R447mil distributable income for the year ended August 2020, no dividend was declared for the year because of the uncertainty created by the Covid crisis. 

South African REIT regulation requires REITS to pay out at least 75% of their distributable income and Dipula has until the end of August 2021 to declare the 2020 dividend. If they fail to do this they lose their REIT status and become subject to capital gains tax. REITS are also taxed (28%) on income that is retained.

Assuming a 100% payout ratio, a dividend of 114c and 54c will be paid to the A and B shares respectively. My impression is that the management team are very motivated to pay this dividend because they do not want to waste the money on tax. It should be possible to pay most of the 2020 dividend. The retained earnings were used to repay revolving debt and this should be easy to reverse.

It is important to note that if the payout ratio is reduced to 75%, the A-share still gets 114c while the B’s dividend falls from 54c to 12c before taking the tax on retained income into account. As stated above, management owns the B-shares so it’s not difficult to imagine why they are so motivated to pay as much of the 2020 dividend as possible.

With respect to the 2H2021 dividend, the payout ratio will depend on ability to fund capex and to repay debt.

Capex Dipula has budgeted R200mil capex over the next 2 years. This will be funded from receipts from vendor loans and R50mil debt. 

Debt Repayment Schedule Debt of R320mil expires in the 2H2021 and a further R1bil in 2022. The probability that the debt will roll is high given the recent willingness of the banks to roll debt at favourable rates. Also, the 100% payout ratio on the 1H2021 dividend  is a strong signal that management is confident that the debt will roll.

Suggested Trades

The two entry points allow investors to mix and match four different trades depending on their risk preference. 

Trade 1: Buy Dipula-A (yield 15%, takes first share of the profit avoiding risk of lower payout ratios or earnings decline, possible 114c from 2020 dividend)

Trade 2: Buy Dipula-B (yield 26%, very sensitive to lower payout ratios and declines in earnings, win up to 54c from 2020 dividend)

Trade 3: Buy equal amounts of A+B (yield 18%, win up to 168c from 2020 dividend)

Trade 4 (Synthetic B-share)

From the B-shareholders perspective the A-share is effectively debt. The B’s capital structure can be simulated by going long the A-share and shorting a bond. (The 10 year bond currently yields 9%).

The synthetic B would involve going long R100 A-shares and shorting R70 bonds. This structure will yield 29% versus 26% on the real B-share.

The synthetic is probably the best risk adjusted trade. Using monthly returns this strategy has a 9% standard deviation which is equal to that of the A-share. The standard deviation on the B-share was much higher at 15%.

The synthetic also avoids the risks of lower payouts or declining earnings which are real risks if another covid wave hits.

Conclusions

A conservative investor should target the 15% yield offered by the A-share. An investor with an appetite for more risk should consider creating a synthetic B at a 29% yield instead of buying the B-share.

Catalyst

- Dividend LDR June,2 2021

- Declaration of 2020 dividend before August 31, 2021

- August results which will demonstrate the resilence of Dipula's income despite the covid crisis.