“One of the hidden misconceptions or something that I didn't appreciate was a n important factor, maybe it's obvious to a lot of folks, but to me at least, as a city boy, just doesn't have that. There's a small dynamic of your high-value add agritech. So you're looking for vertical farms, some of your agri-tech high-level productivity. Somewhat next to it are food processing, and food trading. So I think when we look at that also, we need to think about the layers, not just in terms of agriculture, in terms of just farming, but before that, you have the inputs, the actual farming, the harvest, you have the offtake in terms of movement. You have to do the processing, then you do some more movement, then you're doing trading, and eventually, you're selling this and guaranteeing shipment. So there's an interesting value chain that follows the processing chain of it. So the goal is just to put up a landscape of that at level one in terms of Agritech.” - Jeremy Au
“There's this dynamic about thoughtfulness. I hope that as a founder, you understand these cash crops slash farms last region. I think there isn't a lot of margin of error for you to be a city boy, to learn this from scratch, deploy and get it wrong because, it's capital. If it's asset-heavy, you had to be on the ground. It's hard to recover from a mistake in every tech if you don't understand what's going on. I respect a lot of people who already have experience in their families because they grew up on those farms. They just have more understanding of what's actually going on. There's honestly a certain level of domain expertise that is hard to learn with your first startup because that feels like a very fragile dynamic to be in.” - Jeremy Au
“One interesting aspect that is underappreciated is that their supposed price is higher, so obviously, you take more inputs because it's a fundamental energy transfer problem. Animals always require more energy because of kilojoules to farm and raise and grow compared to plants. I'm saying that there's also an interesting dynamic where people are making an argument that as countries get richer, for example, in Southeast Asia, their demand for protein goes up. So chicken tends to go up, or fish would be a big one as well. So there's different dynamics between input price and then the sum total of that, versus the processing, versus the eventual retail price.” - Jeremy Au
1. Southeast Asia Agriculture: Jeremy and Adriel talked about the variety of agriculture players, from smallholders to large conglomerates, and the range of crops cultivated, including staples like rice and cash crops like palm oil. Jeremy pointed out that understanding agriculture requires consideration of geography, market players, and crop types, noting that this sector is significantly more complex and boots-on-the-ground vs. well-covered business models like SaaS. They explore upstream inputs, financing, downstream off-take, storage, processing and trading functions. They also discuss the complexity of determining investment profiles suitable for venture capital return targets.
2. Digitization & Mechanization: Jeremy compared the region’s agricultural landscape to the US, where both hardware and software advancements have been implemented. He explained that despite the potential benefits of digitization, the fundamental nature of agriculture requires a physical interaction with crops, which cannot be entirely replaced through digital solutions, and stressed the importance of understanding the practical limitations of agritech. He also discussed the potential of logistics as a fundamental layer in the agriculture sector and pointed out that while there are opportunities for increased efficiency through better maintenance, routing, and time management, the gains might still be limited.
3. Livestock vs. Crop Sector: Jeremy delved into the differences in managing poultry and livestock compared to crop-based agriculture. He explained that different animals require varied inputs and are subject to different risks, and recognized the higher price points for animal products due to their higher energy input requirements compared to plants. He also emphasized the anticipated increase in demand for protein-rich foods as the region grows richer. He discussed the impact of diseases on poultry and fishery yields and shared how to manage these risks, comparing it to an insurance model where consistent monitoring and preventative measures are vital.
They also talked about the role of government support in agriculture, the impact of weather and seasonality on agricultural productivity & asset utilization, and the potential of cultured proteins.
Supported by Fluid
Did you know that over 70% of B2B trades are conducted on credit terms? Yet, many suppliers struggle to support this, leading to lost business opportunities. Fluid offers instant B2B financing with one tap, seamlessly integrating with marketplaces and supplier platforms. This payment flexibility empowers buyers to secure their purchases on credit terms or installments. This results in increased basket sizes and an influx of new buyers for suppliers. Fluid provides a great user experience and the ability to facilitate high-velocity trade. This differentiates Fluid from traditional digital lenders and invoice financing companies. Want to learn more? Get in touch with Trasy, Fluid's cofounder, at firstname.lastname@example.org to learn more.
(02:01) Jeremy Au:
Hey, Adriel, what's up?
(02:02) Adriel Yong:
Hey Jeremy I guess it's time for our monthly recording where we riff about different things we have seen across Southeast Asia. I think we have looked at a lot of agricultural stuff and discussed agricultural companies at quite a deep level over multiple occasions, and I thought it would be a good opportunity to just synthesize some of our learnings and thoughts whether it's in agriculture in Indonesia or even Vietnam.
(02:25) Adriel Yong:
I think a lot of people are looking into e-fishery for X, and agriculture has become one of that X. So what do you think about agriculture technology? I mean, there has been so many different plays across the value chain. You see that in the upstream side where people are trying to do inputs. You see that in financing, in software downstream side where people try to do distribution to retail, F&B. Is this even venture backable? After all, it's a fairly thin margin, fairly volatile commodity as well.
(02:55) Jeremy Au:
Yeah, I think lots of different questions, right?
(02:57) Jeremy Au:
We can start from the, bottom up, we'll work our way up, which is, level one is, how big is agriculture, as a section of the Southeast Asian economy, right? And the answer is, it's large and it's important for both food security, but also for putting food on the tables. So it's for people around Southeast Asia and across the world. We look at, for example, in Asia, Vietnam, there's very large agriculture. You look at the rest of ASEAN, there's also very strong agriculture constituency and sector. And then you, if you look at it, it's not just smallholders but there are also cooperatives, as well as, conglomerates. So there are different kind of market players at different stages. And I think the other slides that we should also think about it is also in terms of crops. So, when we talk about putting food on the table, we're looking all the way from rice, which is your staples, to something that's more on the other end of the scale, which is like cash crops. The most homogenous, nice, for example, would be, palm oil. And then, of course, somewhere in between, you have different types of fruits, ranging from durian plantations to, I mean, there are all kinds of fruits, right? And some of it is domestic consumption and a lot of it is also for export. So it's a mix of that as well. So when we think about agriculture, to some extent we have to think about, again, not just geography, but also what the market play is in terms of the plane of organization that were there, as well as the crop type and the attributes of that crop, which is I think underappreciated because when you're doing SaaS, there's only one type of SaaS, so there aren't too many variables.
But I think crop is, from my perspective, after several years, one of the hidden misconceptions or something that I didn't appreciate was a n important factor in terms of thinking about it, which, maybe it's obvious to a lot of folks, but to me at least, as a city boy, just doesn't have that.
Of course, let's spell out the obvious is that Singapore doesn't really have much of an agritech space. That being said, obviously there's a small dynamic of your high value add agritech. So you're looking for vertical farms, obviously some of your agri tech high-level productivity. Somewhat next to it is some of the food processing, food trading. Obviously, we see Olam doing a lot of that agri-foods commodities movement as well and all the way to agricultural trading as well.
(04:50) Jeremy Au:
So I think when we look at that also, we need to think about the layers, not just in terms of agriculture, in terms of like just farming, if that makes sense, but like before that you have the inputs, you have the actual farming that's happening, you have the harvest, you have the offtake in terms of movement. You have to do the processing, then you do some more movement, then you're doing trading, and eventually, you're selling this and guaranteeing shipment. So I think there's a interesting value chain which follows the processing chain of it. So the goal is just to put up a landscape of that at level one in terms of Agritech.
(05:18) Jeremy Au:
I think very quickly, the second part of Agritech is just talking about what are the opportunities in Southeast Asia, and I think that's something that has a lot of debate, honestly, both bull cases and bear cases and happy to dive more into it as well.
(05:30) Adriel Yong:
Yeah, so I think obviously a lot of people are excited about financing players. And, financing has been done in many different parts and different verticals, right? Not just agriculture. But what do we think about financing for agriculture only? Is that something that's venture backable? Can you underwrite agriculture loans without your loan book exploding in your face?
(05:50) Jeremy Au: I mean, it's hard to tell, right? So I think there are three parts of it. One, you said, is that financing doable? And then secondly, is it venture backable? And the third is how do you avoid errors in the financing side, right? So those are the three dimensions that we're talking about. And again, I want to say that I'm not an expert on this. This the point of view of someone who's still, from my perspective, a beginner, still very much learning about a space as well. So these are the kind of parameters I'm thinking about.
I think first of all, I think that there is obviously requirements for capital. So if you're in agriculture, you're a farmer, you need capital. So you're using a savings and, there's classic development economics, we call the poverty loop, right? Just that, the negative version of that is you don't have enough capital for some whatever reason. You don't plant enough. You can't fertilize enough. Then you don't get good enough harvest, the whole family suffers. Your productivity is lower, you have less capital, and the whole cycle repeats. So the negative version is taught in almost every development economics textbook. And so the conversation is as a government, as a private sector, how do you, how about each, each of them so that, you break out of that poverty loop from an agricultural perspective. So I think capital, whoever for eventually shows up as, but it shows up as subsidized or, stabilized ranging from fertilizer to seeds, whatever it is, to helping support them when there is some sort of seasonality, bad weather so that, a bad climate event doesn't suddenly cause them to be okay for 10 years and suddenly get hammered into poverty in the 11th year. So I think there's a lot of ways that capital can show up, even to guaranteeing the offtake. Obviously it is a huge benefit to farmers because they need a stability, especially for smallholders from plantations, conglomerates. They obviously have their own capital pools to stabilize their own approach in general.
So, I think it's helpful. And I think farmers will always desire more capital. And I think in general, what we see as a result is that most governments around the world effectively has some sort of program, helping of capital, right? And that ranges all the way from government grant programs, to field offices that kind of like buying technology, demonstrating technology, doing local education, all the way to even building out seed banks and different kinds of technologies to increase productivity like high yield crops, high yield rice. So there's a lot of government support and there's also agricultural banks that are effectively doing both grants and low interest loans to even larger loans as doable. And then you see also government support for cooperatives. We're just trying to help these small holders band together and get some level of synergy, across that.
So I think the answer is that in terms of lending, there's both sides of it which is, it is needed. I think there is value and I think it does stabilize that and there's also some level of support. The question that everyone doesn't really know is how much more financing can you do? Because if the answer is a lot of financing can be done, and it can be paid back well, because it's going to be deployed well, productively, then hypothetically, that makes it more likely to be venture-backable. I'm going to put a word venture-backable, converted to high returns versus low returns because I think venture capital returns are much higher, by the way, than private equity returns requirements, which is much higher than commercial return requirements, which is much higher than what the government returns require. So, it's an interesting dynamic where this capital has a large level of marginal social benefit. So every dollar put out is not just commercial benefit, but it's actually a lot of positive societal benefit. So I think there's this interesting dynamic which makes it bit of a head scratcher for me personally, I'm still figuring it out and so I think the pure financing play struggles because even if you deploy, let's say, double the capital, it may not be deployed productively, right? And what that means is that it may not be translated. It may be translated into ineffective, productive techniques. So it doesn't generate the high level return that's needed to justify a VC to be doing this instead of a government program that has a lower return target.
So, there's a dynamic. And by the way, this logic also goes in reverse as well but I think this generates this trend that we're seeing of startups that are focused on a certain cash crop or crop, whatever it is, and saying we need to handle not just financing, but also the deployment of that capital into that set in order to generate productivity. So we're looking at, solar cells, irrigation pumps, fertilizer, inputs. So I think this is dynamic, which is, I think, needed. So you can't just provide capital because it's not going to deploy. Well, I think it will be deployed productively, but may not be at a rate that's high enough for that return profile.
So you're trying to do both. But then of course, now you're starting to get expensive because the point of lending initially was that you wanted to be a very digital, very asset like business. You could do that remotely. You can do remote scoring using satellites or field visits once in a while. And then now you're doing training. And to make sure productivity is happening, you're doing enforcement. And that's where it starts to get heavy in the sense that now you have people, and then, obviously this is geography and so so forth. And then suddenly, you're entering this realm where you're building up and then you're like, if you're doing productivity, then, what other things that are very painful for farmers?
It's they have a lot of middlemen that may be relatively efficient right now from their perspective, but you believe that you can orchestrate a better processing chain that gives better margins to the farmer, but gives yourself some of that margin. And then so you end up following that thesis and you become a one stop shop for that crop, right? And I think that's what we're seeing right now is like almost agri tech is like, they're not just all agri for a country, obviously, because it's a geographic requirement of it, but it's also agri at the whole crop cycle, and they follow that crop from all the way from the start to the end.
And one thing to caveat is, it's not necessarily at the country level, obviously, because crops are really not done at the state level or city level, right? It's really at the farm level. The geo or region, so I think you can say this is an Indonesia chicken startup, but it's not because which island, which region ? It's not Indonesia. It's not pan Indonesia, So I think there's a lot of language that make sense on a digital or lending side, which allows you to have that dynamic to I think what we see are the quirks I think of the agri side.
(11:00) Adriel Yong:
So, I think we talked a bit about software, digitization. What about mechanization of agriculture in Southeast Asia? Which is a better approach? Are farms ready for software or should they, look at more of the drones, the sort of automated robots to replace farmers? What would come first in Southeast Asia, because obviously in the United States, we see sort of both hardware and software.
(11:22) Jeremy Au:
Yeah. I think the awkward reality is that no matter how much software you do online, the seeds got to grow in the soil. It's no matter how much code I write is like that little rice, at the paddy, it needs somebody to take care of it in the soil. So this internet doesn't convert it to some aI robot that, parachutes out of the sky and does the work and then goes back to HQ. I mean, maybe one day, maybe 50 years down the future, we could see some of that mechanization in that format. But I'm just saying right now, the answer is the awkward reality is that there's this physical interface that has to happen. The question is, is it the owner of the asset right now, the farmer, or is it some sort of employee that is for example, of a conglomerate, or is it an employee of yours, or is it some asset of yours, or maybe it's drones, right? But I'm just saying to some extent, there's a physical interface that you just can't really get around for agri, and, I'm just talking about the planting and the harvesting side. And planting has a totally different, if you think about it, dexterity requirement in terms of the motions that you do, harvest is totally different, subject to weather there's so much stuff, variables that, you just cannot assume this is going to be done digitally. And it's kind of funny to say this, but I think it's underappreciated, especially because there's so many VCs who are city boys like me. So, I think you just got to get out to a farm and you got to see it for yourself. You got to see the slaughter house and you got to see, stop being moved around and you'd be like, Oh, this is what's happening. What're the margins you can have out of that business? Despairing supplies to a place, supplying the harvest out to the mill and supplying that stuff and moving that stuff from the mill to your end market. That's a whole trucking business effectively, in many parts of the world.
So I think that's one part of it, and because of that fundamentally is a trucking business, a logistics business, how much margins can you get out of it? The truth is not much, right? I think you look at that value chain and you'd be like, okay, the maximum efficiency we can get out is, let's be real, right, 1 percent more, 2 percent more. a trucking business, right? so I think, versus, financing, like you said, but then what is that percentage on financing ?
I think people, they have a head scratcher because financing is a function of productivity improvement actually, because it's easy to lend money, as we've talked about this in past podcasts with Shiyan and so forth. Lending is not about giving money. It's more like, when you gave the money, did farmer make the right choices? Or did you help them orchestrate that so that there's a high yield in terms of productivity and harvest and so forth? And then that generates the returns for them, so getting paid back the principal and the interest. That's what generates the margin, your net interest margin for this, right? So I think people kind of like forget this stuff for some reason, because I was very excited to land, but not no one's very excited to collect. And then the truth is not performing loans. There's a lot of seasonality, seasonality is hilarious because normally in business world oh, seasonality means, Christmas, everybody buys gifts. Black Friday, or some random stuff but there is literal seasonality seasons for agri tech, right? And it could be a bad season and so, you could have five years of good harvest, and everybody's not performing well. And then you can have flooding, or you could have a hurricane, or tsunami even, right? So there's all kinds of weather events, and then everybody would be, all that risk is correlated because it's based on the weather, right? And so all of them will be non performing at the same time. So there's a huge amount of risk that's there as well.
(14:14) Jeremy Au:
Anyway, I just think that this, all these dynamics that we're not fully sizing the full risk. And then it kind of goes back to your fundamental question is, therefore, as a result, it's like, what is that digital layer that's doable? And I think there's this dynamic where we take a giant step back here to say look, how have other countries increased agricultural productivity time? I think the U. S., it's quite simple. I mean, there's been massive consolidation of all the smallholders. So, most Americans are not farmers, right? So, and if they are farming the farming for different sets of reasons, et cetera, but, most food in America is being built in terms of infrastructure by conglomerates, you can call them but they also service by, providers like Monsanto, right? For example, So these are, acres and acres, they've all been, bought, sold and basically agglomerated into, flat and if the infrastructure all done. So it's all industrialized in terms of the fields in terms of the irrigation, in terms of the wells, in terms of the drones, everything's all simplified to that level. And if you think about it, what I meant was that, years ago, all these individual farms, a hundred years ago, all used to be held by small holders and all of them moved to the city and all of them sold their little farms. And so, that's how it was done.
And then I think you can make an argument. The other side of the table is there are countries around the world where, farmers don't have technically ownership to the land. So they don't have land rights. They just have the rights to use the land, but they don't have the rights to sell the land, And in those countries, you can't sell, and because you can't sell, and then the classic thing is you have two children, you split your farm in two, and the next generation, you split your farm in two again for your two kids. So everybody has a very small farm that can just feed a family at that size, but there is no roll up of all of the land that makes it easy for this straightforward mechanical improvements as well. So then, obviously, the promise then you have is that hopefully there's some technologies that can work in I wouldn't call it fragmented, but, the truth is I think the normal state of the world, lots of smallholders, So, in China, there's like lot of agricultural drone activity, So, instead of trying to deliver some of those services through the normal mechanical route of irrigation lines and so forth can we do that with drones who can get redeployed to spray some fields and then move on to the other field. But it doesn't have to stop in between, right, in that sense.
So it's not that mechanical at the ground level, which requires infrastructure build out. But more on that air level, but definitely not digital. So these are the dynamics that we have to be aware of.
(16:17) Jeremy Au:
Agriculture is a 10 margin. In trucking and logistics, there's also a 10 margin. So the question is what's the productivity improvement they can that 's there. So obviously, you're trying to get more efficiency out. That's the primary way to generate margin because you're trying to get more utilization of your asset. So instead of 70 percent utilization, you're trying to get it up because you have better maintenance, better routing, spending less time. But I think there's obviously a maximum limit to how much you can generate because again, it's a cost center at some level. So I think you're focused on economies of scale, you're focused on routing, and these are a lot of pieces that we see obviously for logistics companies like FedEx. Obviously for UPS, that's your consumer and package side. They're benefiting a lot from the boom in e-commerce, so they'll drive a lot of packages, both on the shipment, but also on the return side. There's a huge volume growth that enabled that economic growth as well, versus, if you think about it if you're thinking about agriculture, let's say you're generating increased productivity of 2% to 5% to 10%. There's not a dramatic increase in load in that sense. So I think there's a conversation which is, are you going to continue working with third party providers, or are you going to move to a first party provider because you can't really do it well? But then again, the seasonality. So very few goods like maybe chickens could be year-round because you have a lot of control around the environment, but if you're talking about rice seasons, again, your assets are going to be unutilized from many parts of the year because the rice is just growing and they haven't been harvested yet.
So I think there's this dynamic about the thoughtfulness. And I think that's where I'd maybe a good segway is I hope that as a founder, you really understand these cash crops slash farms last region. I think there isn't a lot of margin of error for you to be a city boy, to learn this from scratch, deploy and get it wrong because, it's capital. If it's asset heavy, you had to be on the ground. I think it's hard to recover from a mistake in every tech if you don't really understand what's going on, versus, I think, I really respect a lot of people who already have experience their families because they grew up on those farms. They just have more understanding of what's actually going on. There's honestly a certain level of domain expertise that is hard to learn with your first startup because that feels like a very fragile dynamic to be in.
(18:08) Adriel Yong:
So, we spoke a lot about rice and vegetables so far. What do we think about things that actually move, like poultry, whether it's chickens, cows, pork, or even fish.
(18:18) Jeremy Au:
Yeah that's a super fair point, which is, obviously there are different types of agriculture slash domain in that bucket, right? So you have beef, pork, chicken, so and so forth, that's across the region as well. And obviously, the requirements are different in terms of what inputs are required, what are the dynamics, what're the risks involved. And I think, obviously the market players, one thing that I've noticed a little bit is that I think there's an interesting dynamic where you have market players already controlling the inputs for this type of dynamic. So, for example for chickens, you have a DOC, which is the old chicks, which is, important requirement, which is what are young chickens that are being born so they can put them the growing facilities all the way to different types of feed that you require So I think there's an interesting dynamic where I think there's a little bit more processing on the inputs, I would say, in terms of, because the animals require different things. I think adjacent to this, you also see some different approaches as well. There's been folks for example, who are working on like antibiotics, and hormones to like more ethical versions of antibiotics. So, there's a startup in Thailand that's building like phage bacteria. So the thesis is that you can use bacteriophages to attack a lot of these diseases without having to use antibiotics. So, interesting approach. So I think there's a different set of considerations because it's just a different crop on that perspective.
(19:36) Jeremy Au:
I would say one interesting aspect that is maybe underappreciated, is that their supposed price is a higher price, so obviously you take more inputs because, it's a fundamental energy transfer problem. Animals always require more energy because of kilojoules to farm and raise and grow compared to plants. But I'm just saying that there's also interesting dynamic where people are making an argument that as countries get richer, for example, in Southeast Asia, then their demand for protein goes up. So, chicken tends to go up, or fish, I think would be a big one as well. So there's different dynamics between, what kind of input price and then the sum total of that, versus the processing of that, versus the eventual retail price.
For example, I remember I went to Oregon and they have the Oregon Truffle Festival. One of the products they're busy looking to farm and try to farm more of is truffles, and because they have a lot of wine plantations and they took on those truffles from Europe, and then they use that to kind of culture and basically seed truffles across their vineyards. What that means is that they had, truffles have a very expensive retail price, and then they also had a tourism festival. So you're never going to see a tourism festival for rice, but for truffles, as a whole, one month celebration of the truffle season where people fly in. My wife and I, we had romantic getaways there and we just enjoy two weeks in wine country with truffles. So I'm just giving an example of how a crop harvest that's has a different high retail price can generate a different dynamic. Same thing for wine, for example. So I think we just have to be thoughtful about that.
(20:55) Adriel Yong:
Cool. And adjacent to that is sort of disease testing because such a large part of agriculture, poultry, fish, the yield is destroyed each time by a sudden disease. You fish flu. I mean, especially for those in aquaculture. It's so easy for diseases to be transmitted in the open sea or ocean. Is that an interesting category that more people should pay attention to instead of trying to put IOTs in farms and all that?
(21:20) Jeremy Au:
I think the question is, how do you capture value in that? So I think the answer is, there's obviously a lot of value it's like a spiky risk. If it happens every 10 years, but it causes a destruction of effectively 100 percent of value, then if you amortize that cost or risk, it's basically creating 10 percent more margin. You can call it right or upside every year for 10 years, but you gotta be there for 10 years. It's like selling insurance. Nobody wants to buy insurance for that moment. I mean, once you need it, then you're like, oh, I wish I bought insurance, but it's too late. So, I don't think a company just focused on that tail risk can survive the time for that tail risk to come in, and it's too late anyway because you need to do, like you said, the water testing, or the, preventative measures, or the monitoring of the soil slash water slash, health to be able to catch that early enough to avoid that happening. So I think basically this is a actually a very important and underappreciated dynamic that's very socially beneficial.
And I just think that it just has to become one of the features of one stop shop that's there, the supporting the other blades of monetization that's there.
To wrap things up, I think one thing, of course, is interesting is that alternative proteins are going after the space as well in terms of all the proteins. And what's interesting about that dynamic is that they're saying, Hey, we can use vegetables and we can use culture dynamics to basically go after these high value protein products. The dream that they have is that, right now, it's more expensive in general. But for me, I do believe that it's a bit similar to the solar cell industry, which is that the truth is there's a fundamental law to, for the conventional farming dynamic to get the cost of beef. But if you're culturing beef cells or so forth, you can really get the cost of production to an order of magnitude lower over the course of maybe not in the next 10 years, maybe over the course of the next 50 years, and then once that happens, once it starts getting cheaper, and it's roughly equivalent in terms of taste.
And similar to solar cell, then it would become a takeover in that sense. But then, maybe there's a story for another day. But I think the question then is if you believe that similar to solar cell industry, then does it reward players who are less about R&D, but more about economies of scale and the lowest cost of production, just similar to how China took over and replaced Germany as the number one exporter of solar cells. So I think there's an interesting dynamic for alternative proteins. But I do think that alternative proteins does has a way eventually get there to be cheaper than the cost of farm protein. And I think that's interesting disruption, you can call it, but interesting future risk point of view.
On that note, thank you so much Adriel. It was fun chatting with you.