Grupo Aval Acciones Y Valores S.A. Q3 FY2020 Earnings Call

Welcome to Grupo Aval's Third Quarter 2020 Consolidated Results Conference Call. My name is Silvia. I'm going to be your Operator for today’s call. Grupo Aval Acciones y Valores S.A., Grupo Aval, is an issuer of securities in Colombia and in the United States. As such, it is subject to compliance with securities regulation in Colombia and applicable U.S. securities regulation. Grupo Aval is also subject to the inspection and supervision of the Superintendency of Finance as holding company of the Aval financial conglomerate. The consolidated financial information included in this document is presented in accordance with IFRS as currently issued by the IASB.

Good morning and thank you all for joining our third quarter 2020 conference call. I hope that all of you and your families are keeping healthy. During the third quarter, several of the countries where we operate started to ease their mandatory quarantine and the effect on their economies was felt almost immediately, more so after the sharp contraction experienced during the second quarter. Our subsidiaries have also benefited, as evidenced by the pickup in banking fees and revenues from infrastructure. We are, however, months, if not years, away from returning to business as usual as economic recovery and health challenges remain ahead of us. In today's presentation, I will cover the following: a macro review of the economy during the third quarter of 2020, a quick update of our loan relief programs, the progress of our digital efforts, and the main highlights of our financial performance. I will spare you from Ruta del Sol-related topics because there have been no new developments since we last spoke.

Thank you, Luis Carlos. I will now move to our consolidated results that were filed under IFRS. Grupo Aval's third quarter results reflect the positive impacts of the economic recovery as lockdowns ended, boosting non-financial sector and fee income performance. In addition, falling interest rates allowed our banks to improve their trading income and realize OCI gains on fixed income. However, the cost of risk remained high, and asset quality showed evidence of anticipated deterioration resulting from leases expiring, which is one of the negative impacts of the economic cycle on a customer’s payment behavior. Starting on page eight, our quality asset growth was 0.3%, while 12-month growth reached 21.9%. As mentioned in our last call, the MFG acquisition was closed last May. During the acquisition of MFG and FX movements of our Central American operations, total assets grew 11% over 12 months. Our Colombian assets decreased 0.6% during the quarter and grew 12.5% year-on-year. Our Central American assets recorded a 1% quarterly decrease in dollar terms and a 28.5 year-on-year growth. MFG concluded with 20.4% of year-on-year growth. Depreciation of 11.2% for 12 months and a 2.9% for three months takes our annual and quarterly growths in pesos of Central America to 42.8% and 1.9% respectively. The rate of Central America increased slightly during the quarter to 36% of our book. Moving to page nine, our loan book grew slightly over the quarter, reflecting an improvement in dynamics of retail lending in Colombia while incorporating the charge-off of retail loans. Central American operations maintained stringent underwriting thresholds with loans contracting in dollar terms and growing 2.3% when translated into Colombian pesos. Loans increased 0.4% on the quarter, reaching a 17.2% growth year-on-year. The acquisition of MFG added COP13.1 trillion or $3.4 billion to our year-on-year loan portfolio growth and explains 7.3 percentage points of 12 month consolidated growth in peso terms. The Colombian gross loan portfolio increased 7.9% over the year and contracted 0.7% during the quarter. This quarter’s performance reflected a recovery of our Colombian consumer portfolio dynamics, a reduction in our commercial portfolio, and the write-off of loans sold. Demand for consumer loans in Colombia resulted in a 2.1% growth in the quarter and 6.5% year-on-year. This quarterly growth was driven by secured products. We also reported an improvement in economic activity and the increased effectiveness of our sales network at the same time restrictions and lockdowns were progressively lifted. Payroll lending accounts for 57% of our Colombian consumer portfolio, which grew 4.3%, and auto financing, accounting for 7%, grew 0.3%. As always, secured retail products, such as mortgages, remain dynamic in Colombia, expanding 3% over the quarter and 12.2% year-on-year. In contrast, credit cards contracted 2% and personal loans decreased 0.7%. These products account for 14% and 22% of our Colombian consumer portfolio respectively. On the other hand, the Colombian corporate loan portfolio decreased by 2.5% over the quarter and approximately 8.2% over 12 months when excluding repos. Corporate lending retrenched to 2.1% over the quarter, mainly among large corporates as they recreate working capital loans disbursed as a safeguard during the first set of lockdowns back in March. Also, as part of our personal growth, we offered to pass on some large corporate and institutional loans that may have been affected by certain market conditions as competition intensified this quarter. Finally, we charged off, we probably sold for COP824 billion that accounted for 99 basis points of poorly placed commercial loans. Moving to Central America, our closed loan portfolio contracted 0.6% over the quarter and increased 23.2% year-on-year in dollar terms. During the impact of the acquisition of MFG, Central America grew 2.6% year-on-year. Quarterly performance resulted from a 1% and 0.6% contraction in commercial and consumer loans respectively. Mortgages increased by 0.3% over the quarter. The performance in Central America's commercial loans reflected similar trends as in Colombia. Consumer loans were impacted by a 0.7% contraction in credit cards and a 2.6% decrease in personal installment loans. We expect that commercial loans growth will continue to be affected by the pricing competition and the leverage of large corporates, both from high-quality institutional and corporate customers. On the other hand, we expect that vehicle loans will continue to recover as the employment outlook improves and households slowly regain confidence in a scenario with no further countrywide lockdowns. On pages 10 and 11, we present several loan portfolio quality ratios. As anticipated, delinquency metrics deteriorated through the quarter, the effect of normal lease expirations and a portion of them transitioning into 30 and 90-day PDLs. As of September, 70% of our loan portfolio had activities. Activities were 14% in Colombia and 22% in Central America. Our loan portfolio deteriorated by 111 basis points, 5.16% on 30-day PDL basis, and 5.1 basis points to 3.21 on the 90-day basis over the quarter. Our 30-day PDLs are now 61 basis points deteriorated throughout the quarter a year earlier. On our 90-day PDL ratios with 5 basis points better. Breaking these figures down by performance; the commercial loan portfolio deteriorated 61 basis points to 4.7% on a 30-day basis over the quarter, and by 13 basis points to 3.7 on a 90-day basis. In Colombia, commercial PDLs deteriorated 61 basis points to 5.9% on a 30-day basis, and remain stable on a 90-day basis at 4.7%, but that was charged up at a 95 basis points, which had a positive effect on these ratios. In Central America, 30-day PDLs deteriorated 74 basis points to 2.1%, while 90-day commercial PDLs deteriorated 55 basis points to 1.6%. Looking at our consumer portfolio, this portfolio deteriorated 188 basis points to 5.7% on a 30-day basis over the quarter and 43 basis points to 2.4% on a 90-day PDL basis. In Colombia, 30-day PDLs deteriorated 152 basis points to 5.8%, while 90-day consumer PDLs deteriorated 7 basis points to 2.7%. In Central America, 30-day PDLs deteriorated 245 basis points to 5% while 90-day consumer PDLs deteriorated by 100 basis points to 1.9%. Mortgage PDLs deteriorated by 124 basis points on a 30-day basis to 5.7% and 15 basis points on a 90-day basis to 2.9%. Our provision expenses remain high during the quarter, reflecting offsetting trends between an improving macroeconomic outlook in Colombia and a softer one in Central America, as well as the deterioration of asset quality measured by stages under IFRS 9. Contractions in riskier portfolios, in our corporate loans favored our quarterly cost of risk. Cost of risk improved by 18 basis points from the quarter, six basis points of which are explained by better recoveries of charged-up assets. The cost of risk in commercial loans improved 82 basis points or that of retail loans deteriorated 63 basis points. The cost of risk in Colombia improved 60 basis points and deteriorated 57 basis points in Central America. The increase in cost of risk in Central America is offset from a pickup in cost of risk for retail loans, partially explained by a recovery in dynamics of credit cards. Our third quarter cost of risk incorporates a 0.7% contraction in credit cards that favored when it compares to a 4.6% contraction recorded during the second quarter. In addition, Central American cost of risk reflects an earlier expiration of reliefs, particularly in Costa Rica and Guatemala. Meanwhile, in Colombia, cost of risk for retail loans was lower during the quarter, mainly due to an improving macro scenario and an increase in recoveries of charged-off retail loans. Our banks continued to calculate with loans according to risk level in order to evaluate expected credit losses and determine impairment charges. This resulted in an increase in stage two exposures and in the Aval Affairs commercial loans under stage three. As mentioned in previous calls, consolidated exposure to the Avianca Group is approximately $100 million equivalent to COP727 billion as of September 30, 2020; 23% of this exposure is secured with international billings and 20% is secured with our Avianca's headquarters buildings in Bogota. Our coverage for Avianca reached 35% as of the end of September. Recoveries of charged-off assets were lower during the quarter, as collection efforts regained traction with the lifting of lockdowns. We expect our provisioning expenses will increase in the fourth quarter, as we incorporate our iteration in Central America and macro expectations. Most related provisions are booked, and reliefs continue expiring, and a portion of stage 2 loans transition to stage 3. Finally, our PDL coverage of 90-day PDLs slightly decreased to one-and-a-half times. On page 12, we present funding and the positive evolution of funding growth during the quarter, which continues to reflect a conservative liquidity profile to face the risk associated with the pandemic. As a result, our deposits to loan ratio increased to 107%. While our cash deposits ended the quarter at 18%. Funding structure shifted slightly toward deposits now accounting for 77% of total funding. Deposits increased 2.3% during the quarter and 24.7% year-on-year. In Colombia, deposits grew 2.7% in the quarter, and in Central America, they grew 1.9% in dollar terms during the quarter. In the 12-month period, Colombia grew at 12.7% and Central America at 34.9% in dollar terms, with 18.9% explained by MFG. On page 13, we present the evolution of our total capitalization, our attributable shareholders' equity, and the capital adequacy ratio of our banks. Total equity grew 9.5% year-on-year, while our attributable equity increased 7.8%, mainly driven by our earnings. Payroll loans were at 3.7% and 4.4% respectively. As of the third quarter of 2020, our banks show a profit Q1 and problem solvency ratios. The quarterly increase in Q1 for Banco Popular is explained by the risk-weighted assets contraction and problem solvency remain flat. We expect that the transition to Basel III and the solvency ratios of our regions will increase a few percentage points while Bogota should remain at a similar level to the current ratio. On page 14, we present our yielding loans, cost of funds, spread, and net interest margin. Our net interest margin performance during the quarter was driven by a stable NIM on loans and a lower NIM on investments. NIM on loans remained stable during the quarter, mainly due to an aggressive strategy to incorporate the Central Bank dynamics that yielded a 31 basis points reduction in the cost of funds. In addition, our second quarter included a COP34 billion reduction in interest income or a 14 basis points reduction in net interest margin on loans associated with the recognition of our increasing present values of loans due to the terms under which reliefs were granted. We expect that competitive pressure on pricing of loans, particularly that for the highest quality risk, will persist as the macro outlook improves. However, when compared to 2020, 2021 is expected to benefit from lower rates paid by our banks, as they fully incorporate the Central Bank intervention rate cuts observed this year. Net interest margin on investments will continue to depend on global liquidity and geopolitical events for the last quarter of the year. On page 15, we represent net fees and other income. Gross fee income for the quarter reflects the increasing activity due to the end of lockdowns and the temporary waivers on transactional and other fees. Pension fund fees were positively affected by increasing performance-related fees on mandatory pension funds and assets under management fees charged on severance funds. Quarterly gross fees increased 14.3% in Colombia and 16.8% in dollar terms in Central America. We expect to see further recovery on this front as economic activity continues to improve over the following quarters. The performance of the non-financial sector reflects the recovery of infrastructure and gas sectors, which are our two main multinational businesses. Most restrictions on the infrastructure sector in Colombia were lifted earlier in the third quarter, allowing construction progress to pick up to pre-COVID levels. The energy and gas sector was positively impacted by a recovery in demand for industrial gas. Finally, on the bottom of the page, higher other income during the third quarter is mainly explained by strong results in OCI evaluation and high income over the quarter. On page 16, we present some efficiency ratios. All of our business units continue implementing cost containment and reduction initiatives during the quarter. As a result, cost to asset improved to 3.1%, down from 3.2% in the previous quarter and 3.9% a year earlier. Cost to income decreased during the quarter to 44%, reflecting a recovery of fees and income for the non-financial sector, as well as strong other income as described before. Other expenses remained flat both over the quarter and year-on-year, despite the acquisition of MFG and FX operations. Excluding these effects, other expenses contracted 0.5% in the quarter and 8.2% year-on-year. Quarterly Colombian other expenses decreased by 3.5% during the quarter and 9.1% year-on-year, and Central American expenses increased by 8.1% over the quarter and 2% year-on-year in dollar terms, now incorporating a full quarter of MFG up from one month in the second quarter. Excluding MFG, Central America other expenses grew 4% over the quarter and decreased 6.9% year-on-year in dollar terms. Our liquid loss in Central America was influenced by $8 million in increasing severance payments. Our consolidated quarterly personnel expenses increased 4.5% over the quarter and 7.1% year-on-year. Over the quarter, personnel expenses increased 3.7% in Colombia and 8.5% in dollar terms in Central America, now including a full quarter of MFG. Personnel expenses grew 1.8% without the effect of MFG, including the severance payments mentioned before. Excluding MFG and FX fluctuations, personnel expenses increased 2.8% over the quarter and decreased 2.2% year-on-year. Quarterly general and administrative expenses increased 2.9% over the quarter and decreased 6.8% year-on-year, collateral reaches a 12.8% decrease when excluding the effect of MFG and FX fluctuations. Finally, on page 17, we present our net income and profitability ratios. Attributable net income for the third quarter 2020 was COP691 billion or COP0.31 per share. Year-to-date attributable net income reached COP1.7 trillion or COP0.77 per share. Our return on average assets and return on average equity for the quarter were 1.5% and 15.6%. Year-to-date return on average assets and return on average equity reached 1.4% and 11.4% respectively. Finally, even though a high level of uncertainty persists, I will summarize our expectations for 2020 and 2021. We expect loan growth to be either 17% for 2020 and 9.5% to 10% for next year. Our net interest margin should remain fairly stable at 5.9% for both years. Return on equity for this year should be close to 10% and improve to 12.5% to 13% in 2021. We will have a 2020 cost of risk at around 3% and next year closing at around 2.5%. Fees should grow a couple of percentage points faster than the loan portfolio, and finally, we will have a 2020 expense growth in the 4% range as we have this year. We'll now open it for your questions.

Thank you. We will now begin the question-and-answer session. Our first question comes from Jason Mollin from Scotiabank.

Yes, hi. Thank you for the opportunity to ask questions. My question is on the level of provisioning, and provisioning is high relative to past levels. You talk about cost of risk potentially around 3% this year, declining to 2.5% next year. One question is how should we feel about the level being sufficient that you have now, especially relative to some of your peers who are making higher provisions if you want to measure it as a percentage of NPLs, which is complicated at this point because if they’re rescheduling, it's hard to look at that ratio in my view, also as a percentage of net interest income or as a percentage of total loans. How can you give investors assurance that the provisioning levels you have now are sufficient?

Thank you, Jason. I will try to summarize some of the things we've said in the past that have become apparent throughout this cycle. We have a different structure of our loan portfolio, where we are overweighted in some of the safest products and underweighted in some products and segments that are riskier. We have a higher portion of payroll loans to government employees and retirees, and we also have a smaller portion of unsecured consumer lending as well as we are underweighted in the SME segment. Part of what you’re seeing at this point is the way our portfolio is evolving and how the different portfolios have behaved. We’re not covering this in the call, but when you look at a micro level, there's a wide variation in the sort of provisions needed for the different products and segments that I mentioned before. As you might have seen, there's an implied slight pickup in provisioning during the fourth quarter to be able to get to 3%, and what we’re doing at that point is raising some of the provisions we have in Central America as well as there's a process of stage 2 loans migrating into stage 3 that should also imply additional provisions. Finally, something we’ve mentioned in the past that is changing is the speed at which we’re growing. Part of the lower provisions previously was due to shrinking our riskier portfolios, particularly the credit card portfolios during the second quarter.

The next question comes from Gabriel Nobrega from Citigroup.

Hi, everyone. Good morning and thank you for the opportunity to ask questions. I actually have two questions. The first is on asset quality. We have noticed that this quarter, but we still saw you have 90 days NPL ratio increasing. I wanted to understand if the 20 bps increase in the quarter is mainly due to Avianca? And then, for my second question, it's actually on the NIMs. You were guiding that NIMs are going to remain stable through the end of the year and also in 2021, but as your stage 3 loan portfolio starts increasing, we have to remember those loans also don’t accrue interest rates, and there's another important part that your banks are going to start re-pricing their loans for the lower interest rate environment. I wanted to understand where you are trying to work harder to maintain net interest margin?

Okay. Let me try to answer your question on asset quality, and if I didn’t get your question right, please correct me. What's going on now is that we've begun to see the reliefs, where the calendar starts ticking first for the 30-day past due loans, and then migrates to higher 90-day PDLs. Historically, a kind of transition that we’ve had from 30-day past due loans to 90-day past due loans could be somewhere around 60% to 70% or something of that order. So, it's a matter of time for some of these to migrate into the 90-day past due category, and that's the reason why we’ve also guided into higher provisions for the fourth quarter. Regarding where things should end, we are fairly comfortable that at this point, we have more visibility and, in the absence of new countrywide lockdowns, we should be evolving as I described previously. Now, moving into your question on net interest margin, there are multiple factors involved. We had a big hit from re-pricing our loans in advance to re-pricing of our deposits before. We’ve already been able to catch up, and there's some room for additional reduction in the cost of funds that will continue helping us. On the other hand, we had, I would say, a temporary effect of reliefs that we’re reducing our net interest margin as conditions in the present value were reducing the value of those loans and we had to pass that through to our interest income. Additionally, more structurally, during this month, we have prioritized growing in very high-quality loans. Therefore, when you see our loans that are somewhat flattish during this quarter, what you have is a combination of shrinking riskier loans and moving our portfolio to higher quality loans.

Diego, very complete answer. Just one thing I would add is to address your question on why the 90-day past due loans had increased by 20 bps, and it is, as you assumed, mainly due to Avianca.

Our next question comes from Yuri Fernandes from J.P. Morgan.

Thank you, Diego, Luis Carlos, good morning. Very quickly here on the cost of risk, if I got it correctly, cost of risk should be around 3% by year-end. This basically implies that for the first Q we should see the cost of risk around 3.5%-3.6%, as you said on page three. My second question is regarding a second wave. We are seeing some countries getting more concerned about that, like hospital utilization moving in Brazil, and we discussed this in Europe and in the U.S. What about Colombia? What is the outlook here for COVID? Do you think if a vaccine is approved globally, Colombia will be able to get the vaccine quickly? There are concerns that for developing countries, it is not clear how quickly they would get access compared to developed nations. So, I am concerned that 2021 could imply a second wave of further lockdowns.

Okay, Yuri, regarding your numbers on cost of risk, you're right, that’s what is implied for fourth quarter. As for the second one, I'm sorry, but I think we're not qualified to opine on the health side. We see differences in the initial cycle of what happened in Europe compared to what happened in Colombia. At this point, it’s a combination of what we hear from public policy, where public policy is trying to defend employment and activity, and then on the health side, I'm really unqualified to answer your question.

We're obviously looking with optimism at the results of the COVID vaccine tests and are hopeful that smaller countries, the developing countries, will have access to the vaccines almost as quickly as the developed countries, which might be the answer to not having to go through additional lockdowns.

At least not countrywide lockdowns.

Our next question comes from an unidentified analyst.

Hello, everyone. Thank you for allowing me to ask a question. I have a question regarding the provision and the risk associated with the loan book. If I understand correctly, this indicates a 3.5% cost of risk in the fourth quarter, which suggests that the highest provision expenses should occur in the first quarter. If that is accurate, could you explain why these provisions weren't made earlier? According to IFRS 9, provisions should be anticipated based on those macro assumptions, like GDP projections. Also, what GDP assumptions did you use for 2020 to estimate the provision expenses based on expected losses?

Let me try to take it piece by piece. Regarding provisioning and timing, I mentioned that there are forces happening here. Some of those are things that we are starting to know now, and some things that have not yet materialized but we feel could evolve as expected. Cost of risk will have a combination of several factors. The macro outlook is one of the components of what goes into IFRS 9, but it is not the full picture, and you have to bear in mind that it’s actually forward-looking. At this point, when we end the fourth quarter, we're actually looking into 12 months forward, which is the full 2021, and that's where you're talking of around 4% GDP growth. So, what I'm saying is, it makes sense that this year would have been, I would say, horrible in Colombian standards. Most of that has already happened; therefore it doesn’t build into IFRS 9. Regarding why we hadn’t provisioned before, I would say because the performance, as we were looking at, indicated that the adequate level of provisioning was being met. I actually provided a cue for where it will come from, and part of what I said is it’s coming from Central America, where the initial forecasts from most analysts were much more optimistic than what we’re currently looking at as a consensus. That counts into why our numbers look the way they do. Also, I mentioned some things that are good quality provisioning, and provisions linked to moving from shrinking some of our portfolios into starting to grow.

He wanted to know about our GDP predictions.

Yes, I think I mentioned that by the end of this year, our GDP projections for next year would be in the 4% level, that's what we're looking into, and that's what's going to build into the IFRS numbers of year-end.

Right, 6.5% contraction this year in Colombia.

Our next question comes from Andres Soto from Santander.

Good morning, Luis Carlos, Diego. Thank you for the presentation. I have a few questions. The first one related to expenses. If I heard correctly, your expenses in Colombia declined 9% on an annual basis. I would like to understand what measures you have implemented to achieve such an important result. The second question is regarding your guidance for loan growth next year. I am trying to understand how this breaks down between countries, in general, between Colombia and Central America, and also within segments? Finally, when I look at the numbers for 2022, I assume that for 2022 you guys will be having a cost of risk already below the level you had in 2019, and with that, the ROE that you should be getting is in the area of 15%. I would like to confirm if that's the number and, more importantly, if this is what you believe is the potential ROE for Grupo Aval or if there is any potential for an additional increase in this number.

I think there are a lot of questions, and I will try to go through them briefly. Regarding expenses, I think something you mentioned we should have highlighted before because we have a very positive year-on-year comparison when we take the third quarter. Last year, during this time, we had some benefits from a tax program that ended up generating benefits on the expense side. However, for the sake of completeness when we take OpEx year-to-date to try to strip those things out, we are contracting something short of 1% is what’s happening. That gives you a better idea, and there was some seasonality there particularly last year, which is a relevant point that you brought up. Then, regarding segment and country growth, what we expect to see is Central America growing slightly on the lower side of the range we mentioned, something in the order of 9% to 9.5% in dollar terms, while Colombia should be growing faster. What will be happening with Colombia is that we had shutdowns during this year for corporate growth. As that growth is coming back, what we are looking at for Colombia is something more in the 10% area. So, it’s basically taking the range with Central America growing at the lower side and Colombia growing at the higher side of that spectrum. Inside Colombia, we should start to see corporate loans recovering some of the room that they have let go throughout the second half of the year. Then moving into cost of risk for 2019, I have to go back into history. When you look at our numbers, our numbers have had these three large cases that we have been tracking over many calls that we got done with last year. If you consider what we were looking at before COVID, we should have been around 2%. Two percent implies that we are already done with Electricaribe, Ruta del Sol, and the SITP. The kind of recovery we are looking for is halfway of what we should have been this year, as we are going up to 3%, and next year we are recovering half of that room, concentrating on the narrow part of the year. Regarding ROE potential, we cannot agree with our peers there. The way we look at those numbers is it very much depends on the kind of portfolio each institution has. At least in our case, we think we can aspire to much higher levels than some of our peers have stated.

Our next question comes from Nicolas Riva from Bank of America.

Thanks very much for taking my questions. I have two questions. On the first one, I apologize if you already mentioned this. I joined the call a bit late. It’s on the relief program. You mentioned that the increase in the 20 bps in the NPL ratio this quarter was mainly driven by clients exceeding the relief program. I was wondering what’s the percentage of loans which are still in your relief program as of the end of the quarter, and how many of the regional clients already exited that program, and also, what’s the plan for the remaining loans still in the relief program? Will you be looking to restructure these loans? And then, for the second question at the end of your initial remarks, you provided guidance for next year. I believe you said 12.5% to 13%. If I look at the third quarter, you already made an ROE above that in the third quarter; close to 14% this quarter, and I would imagine that things should improve from here, assuming there are no more lockdowns in Colombia. What is the reason for this guidance, which looks a bit conservative for next year? Is it basically related to cost of risk that you probably expect to still remain high based on this migration of loans to stage 3, or is there anything else going on?

Okay, regarding the relief programs, the number we mentioned in the call was that as of the end of September, we had around 17% of our loan portfolio relieved. This number is more of an order of magnitude, but by the end of October, the estimates would be around 15%. So, we’re seeing a return to normal at that sort of pace, and if you’re thinking of what percentage of the reliefs have expired, I would say around half of those have expired. When you look at the reliefs we've provided, there’s a portion that went back to paying and have already amortized, and there’s a portion that are paying again, and then, within the reliefs, there’s a combination of some of the older reliefs and some more structural ones being given out that will remain for some time after we’re done with this cycle. Regarding the ROE, perhaps to break down what happened this quarter, and why we are guiding into a shire quarter next year, this quarter included the realization of OCI gains on our fixed income portfolio, which should add up to roughly I would say COP130 billion to COP140 billion in a trivial earnings core group to Aval, so that gave us a particularly strong quarter. When we exclude that, if we keep up slightly our provisioning expenses, we would come up with the numbers that we guided. Your deeper question on what's going to happen next year: yes, you're right. At this point, we can't be extremely aggressive with the recovery rate and that's why we are standing by these numbers.

Our next question comes from Sebastian Gallego from Credicorp Capital.

Hi, Good morning everyone. Thanks for the presentation. I have a few questions. The first one is just a follow-up on the provisioning question. I understand your explanation about a more defensive portfolio, but still, you have lower coverage, higher PDLs and higher release compared to peers. Why are you confident that the cost of risk should be at 2.5% next year? Second question is regarding ROE. You mentioned that you don't agree with peers and that it depends on the institution. I would like to know what the outlook for ROE is next year when excluding the non-financial sector. Lastly, my last question would be, at the holding level, we have seen some deterioration in net debt to core earnings, net debt to cash dividends, and double leverage as well. Can you provide an outlook on those measures and if you're comfortable with the levels?

Regarding your question on provision expenses, you have to keep an eye on numerical movement regarding coverage ratios, because coverage ratios are not the same for each one of the products. The provisions needed for a payroll loan are vastly different from those needed for an unsecured consumer loan. The numbers we come up with are based on segment performance analysis, looking at the product nuances I previously mentioned. If you look at our historic comparisons with our peers, this might provide additional context. That is the reason we feel comfortable with our provisions. As for our ROE, it is important to highlight that it is certainly not just the non-financial elements, but also the diversification of our full book. Our performance from Central America has been quite positive, and the countries in which we are operating are also faring better. Therefore, when considering ROE, I would not segment any one component - we have a diversified portfolio; one might contribute better one year, while the following year, another segment turns around. Regarding the holding level, we’re trying to ensure our dollar-denominated bonds are repaid with dollar-denominated assets.

Our next question comes from Carlos Gomez from HSBC.

Thank you very much for the presentation. Two specific questions: one is the level of activity that you are seeing in the fourth quarter, both in Colombia and in Central America. In some countries, we have started to see activity falter after the recovery, and we wanted to know how you are seeing your own numbers internally.

Your question was the levels of GDP?

Level of activity or growth sectors. I'm sorry if it was not clear.

Yes, well, you're right, the levels of activity do vary widely in the regions where we operate. We've already mentioned much about Colombia, but if we consider Central America, the kind of behavior we’ve seen there is that this year, Honduras and El Salvador are the stronger performers, growing roughly around 6%, while Costa Rica and Guatemala are growing closer to 3.5%, and Nicaragua will be contracting roughly 3.5%. Moving into next year, Guatemala and Panama should be growing north of 10%, while Costa Rica, Honduras and El Salvador should be closer to 8%, and Nicaragua should be somewhere flattening to slightly positive. So yes, we're looking into different growth rates when we compare the countries in which we operate.

Our next question comes from Piedad Alessandri from CrediCorp Capital.

Hi, thank you very much for allowing my questions. I had a question regarding the addition to the Pat program, and could you give us a bit more detail on that? Then I wanted to ask, when do you expect to have the peak NPLs for the bank? And finally, what is your exposure to real estate in Central America, specifically in Panama?

Regarding real estate exposure, we likely have the lowest in the market there in relation to size; I can't give you a number, but that's been a business we’ve shied away from from the very beginning, and we’ve reduced those portfolios as we have acquired banks in the past. Concerning the addition to the Pat program, we are in the process of migrating some of our loans there. We are smaller than what our peers have done at this point, but we are indeed starting that much more. Percentage-wise, we could be somewhere around 1-2% of our portfolio that has gone into past programs. Regarding the peak of NPLs, we are watchful of how things should evolve, and we should expect to see NPLs increase during the fourth quarter and remain relatively high during the first half of next year. The numbers should then start to revert to much more normal figures.

Our next question comes from Julián Ausique from Davivienda Corredores.

Hi, everyone, and thanks for taking my question. I have a specific question about the Panama operation because in conversation with your peers, the outlook in this country is a little bit worse than others, so I would like to know what your expectations are in Panama specifically and what is your expectation concerning provisions? Some of the relief funds will end in December or January.

You are right, Panama has been hit strongly. If you're looking into GDP growth expectations for this year, it's estimated that the region could be contracting around 6%, but Panama should be contracting around 10%. The reason for that is that the quarantine in Panama has been much more stringent than that in the rest of the region. However, next year, the regional GDP growth could be somewhere around 3.5%, with Panama growing at around 5%. I would say it’s a combination of how much they got hurt this year compared to what is expected next year, where they may be one of the countries, alongside Honduras, that could grow the fastest in the region. The reason for seeing a different story than others is that we’ve steered away from very risky businesses in Panama, particularly mortgage lending, which we took off our portfolio.

What is the reason that the efficiency rate here in Colombia is too low? I heard something about that, but I couldn't get it.

Yes, I would say it hasn’t been low this year; it was particularly low during the third quarter of last year. That's why we used that to give some information on year-to-date performance.

Do you mean it's good or bad?

Our final question comes from Brian Flores from Citibank.

Hi, thank you. Just a quick follow-up on Avianca; could you reiterate the level of coverage that position has?

35%.

We have no further questions. I will now turn the call over to Mr. Sarmiento for closing remarks.

Thanks again, everyone, for your great questions. We're happy to take more of them if you wish, on a one-to-one basis. You can contact the team through Diego, and in the meantime, we expect to keep it up, and hopefully have you all back in our next quarterly call. Thank you again. I'll see you next time.

Thank you, ladies and gentlemen. This concludes today's conference. Thank you for participating. You may now disconnect.