Description

Turkish equities are offering investors with a multi-year time horizon the opportunity to multi-bag. In $ terms, Turkish equities are down 2/3 from peak. The cheapness today extends to both the currency and the local stock market which, when combined, makes a doubling (in $s) from here for Turkish stocks a very reasonable base case.

How can we frame the cheapness? Lets start with stocks -

• Turkish stocks (MSCI Turkey) trade at 1.08x book today
  • This is the 2nd lowest of any major country in the world (after Russia)
  • This is a 25% discount to the post-GFC median of 1.4x book
    • This discount to history is the largest of any major country that I can find today
• Turkish stocks trade on 6.4x NTM PE
  • This is the 2nd lowest of any major country in the world (also after Russia)
  • This is an 22/12% discount to the post-GFC median/average of 8.3x/7.4x NTM PE (respectively)
    • This discount to history is also the largest of any country that I can find today
• To me, it help to dissagregate the Turkish index into financials (basically banks) and non-financials
• The banks, on average, trade at the following metrics -
  • 3.5x NTM PE
    • Compared to their LT average/median of 7/8x NTM PE (respectively)
  • 0.51x book (they genarated an aggregate RoE of ~ 18% in 2017, and are expected to generate an RoE of ~ 15% this year)
    
    • This compares to a LT average/median of 1.5/1.3x BV (respectively)
    • This also compares to the average large-cap EM bank which trades at ~ 1.9x BV
    • GS just put out a report which uses simple Gordon Growth Model math to suggest that the implied Turkish bank CoE is ~ 25%
  • The operating performance of the banks (to-date) has been remarkably robust
    • NIMs have been expaning for the last 5 quarters
    • Fee income as a % of assets seems to have stabilized after a long period of erosion
    • Cost/Income ratios have been declining for the last 6 quarters
    • NPL ratios show limited signs of the stress implied by Turkish asset prices making their way onto bank balance sheets (yet...)
    • Equity as a % of assets has hovered at a reasonably robust 11% for the last 3 years
    • It's important to note given the FX moves we've seen that the banks are, in the first derivative & in aggregate, net short TRY (i.e. they overhedge their $ asset/liability mismatches) - this claim would be harder for me to accept were it not for the fact that these banks have all spent the last two decade managing/hedging/living with an ever-weakening currency 
• Moving onto the non-financial sector, I think the best way to convey cheapness would be to give you a flavour of some of the value on offer among some of the Turkish corporates -
  • Ford Otomotiv & Tofas are the listed subsidiaries of Ford & Fiat in Turkey (parents each own ~ 40% of the listed companies)
    • They trade at 11.5x & 7.6x 2018 earnings respectively
    • They are going to generate (on consenseus #s) 44% & 37% RoEs respectively this year
    • In aggregate, they have grown revenue (in $s) at a 12% CAGR over the past 15 years (since Erdogan came to power)
    • They've achieved that growth while paying out ~ 60% of PAT in dividends to shareholders
  • Arcelik is a white-goods manufacturer which sells domestically as well as exporting globally
    • The stock trades at 8x 2018 earnings
    • The street has the stock doing a 15% RoE this year, however if you look at the historic correlation between Arcelik RoE and TRY REER (more below), one should expect the company to make a ~ 23% RoE at current FX levels
      • If that correlation continues to hold - the stock is actaully at ~ 5x current earnings power
    • Arcelik has grown revenue (again, in $s) at a 6% CAGR over the past 15 years, while averaging a 15% RoE
  • Sok Marketler seems to be the Dollar General of Turkey
    • The company listed earlier this year (an impressive feat given the Turkish macro/political backdrop) given the parent company's need to delever
    • If you haircut management's store growth plans & haircut street assumptions on LFLs, I still have the comapny generating ~ 22b TRY of revenues in 2020
    • The company disclosures that mature stores (older than 4 years) generate 5.5% EBITDA margins (not crazy - DG/DLTR make > 10% EBITDA margins while large DM grocers make 5-6% EBITDA margins)
    • That would suggests 2020 EBITDA of ~ 1.2b TRY
    • The company will be in a net cash position by YE '18 following the IPO cash infusion + debt paydown - growth in the negative NWC balance more than pays for new store CapEx over the next few years such that, on these high-level numbers the stock is currently trading at ~ 2.5x 2020 EV/EBITDA

Lets move onto the currency -

• On a REER (real effective exchange rate) basis (per the BIS dataset), the Turkish Lira (TRY) is at a 30% discount to its LT REER since Erdogan came to power in 2003
  • There is a pretty robust LT correlation between TRY REER and the Turkish current account (CA) surplus/deficit (which makes sense)
  • The current TRY REER is a level at which Turkey has historically run a CA surplus (vs the ~ 5-6% of GDP deficit its been running over the past year)
  • So, for all the legitimate concern about the funding of Turkey's large CA deficit, the TRY has already moved to a level which effectively (via reducing imports and/or increasing exports) should get the CA to a position that obviates the need for any significant external funding in the medium term
• The hyper-competitiveness of Turkish exporters today further highlights the "cheapness" inherent in the Lira today
  • Touched on above, there's a very high correlation (~ 70% R Squared) between TRY REER and Turkish export-focused comapny RoEs - at the current TRY, they should, in aggregate, be generating roughly 35-40% RoEs vs a historical average of ~ 20%
  • I'm not going to deny that the cost of equity is high in Turkey today but its not 40% high
• The TRY currently offers investors real rates of roughly 4.5% (3m TRY LIBOR is ~ 20% vs last inflation print of ~ 15.5%)
  • This is among the highest globally, after Argentina & Russia
  • If/when TRY volatility declines (which I think is likely in the not-too-distant future given that we're now through elections + cabinet formation + the first post-election CBT decision), I think the perceived attractiveness of the carry trade (both nominal & real) will help to put a floor under the TRY

So why are the Turkish currency & Turkish equities cheap today?

Turkish risk assets have sold off this year given a confluence of negative factors -

1. EM assets started weakening at the end of March this year, from which point EM is down ~ 15% relative to the S&P
2. The underperformance seems to be being attributed to (1) tightening offshore dollar liquidity as the Fed simultaneously raises rates and engages in QT and (2) increased trade war rhetoric, which is perceived to be particularly damaging to EM economices
3. Turkey suffered more than most EMs given -
   • In a time of broad EM weakness, "vulnerable" countries (i.e. CA deficits, relatively limited FX reserves) get hit the hardest
   • Turkey is a major oil importer, so the rally in crude has driven a worsening in Turkey's pro-forma CA position 
   • Erdogan had been running the economy "hot" in order to keep consumer sentiment buoyant going into the June elections, which manifested in accelerating inflation earlier this year
   • Following Erdogan's victory, he appointed his son-in-law to run the FinMin + Treasury & tweaked the laws governing who could be eligible to run the central bank and for how long etc, both of which caused the market to get nervous about the likely quality of fiscal & monetary policy going forward

It's hard to deny the relevance of any of the above - QT is likely bad for offshore $ liquidity, higher oil prices do hurt Turkey, running an economy hot for political purposes is not good policymaking, more experienced/qualified treasury secretaries are generally preferable to less & more rather than less central bank independence is fairly consensually a good thing. So it's hard not to see why the narrative on Turkey is negative today. But lets consider the following. The CA situation should resolve itself over time given the FX adjustment we've now seen. Now that we're through elections & Erdogan is cemented, he should be more inclined to take his foot off the gas and let the economy cool. You can in-fact already see that from the recent banking sector loan growth data. We've already had a decent rate hike in May which has taken real rates to above-historical levels. Given that the CBT just hiked rates in May, I have some sympathy with them seemingly wanting to see what kind of effects that has on the economy before going "full-Argentina" and hiking ad-infinitum to appease the market. Having Erdogan's son-in-law running the FinMin + Treasury could end up being a good thing if at least the country gets someone in charge of the economy who Erdogan listens to & he did spend the last 2.5 years runing the Energy ministry.

So - yes there are negatives with the state of the economy & the economic policymaking apparatus today but I'm not sure they are as bad as some would make out and they appear to be more than fairly discounted by equity & currency valuations today. If the currency went back to its post-Erdogan REER, that would be ~ 45% upside from here. Banks are ~ 1/3 of the equity market - going back to LT BV multiples would suggest ~ 175% upside there. I've looked at the banking sector RoEs over the last 20+ years and there's no obvious downward trend to justify lower PBs than in the past. Corporates are the other ~ 2/3 of the equity market - going back to LT PE multiples here would suggest ~ 30% upside. Putting it all together would suggest ~ 75% upside to the equity market. Combined with the FX upside, you're looking at a combined ~ 150% upside.

If anyone else is following this closely / has a different view of the situation, would love to get a comments section going here.

Catalyst

- Time passing.