Filecoin ESPA - Business Content, 26 May 2022

Previous Meeting

Next Meeting

⏯

youtube image

►

From YouTube: ESPA Module - Filecoin Service Provider Financial Analysis - Acrontech

Description

Tao Tao with the Acrontech discusses their Filecoin Storage Provider Financial Analysis at the Enterprise Storage Provider Accelerator (ESPA) bootcamp week that took place in April 2022.

ESPA is a program built entirely around the use of Filecoin’s Web3 infrastructure. The decentralisation of storage providers and the availability of globally diverse and auditable data is a solution to the misaligned incentives of the server based Web 2.0 storage providers of today. It also sees added benefits of a flexible open spot market for storage and affordable prices set by independent storage providers bidding for the right to house data. All of which are important considerations for enterprise needs and large datasets.
ESPA, an initiative sponsored by PiKNiK, Protocol Labs, Seagate and AMD benefits everyone from the participants taking in new knowledge and know-how to the end users who will have a more efficient, secure, and robust way to store their data.

At the end of the program, participants will be able to leverage the untapped storage space available on Filecoin by onboarding large datasets.

💻 Sign up for upcoming cohorts as a participant:
ESPA: https://www.web3espa.io/?utm_source=y...

🙌 Stay connected with us:
Twitter: https://twitter.com/web3ESPA
LinkedIn: https://www.linkedin.com/company/web3...

💡 Learn more:
PiKNiK: https://www.piknik.com/?utm_source=yo...
Protocol Labs: https://protocol.ai

A

All right so um appreciate it, everyone, sorry, I cannot make the trip to vegas. um I was traveling last week, so I got too much going on at home, but I appreciated this opportunity to be able to talk to you guys virtually and nice to meet. You see you guys again the picnic team and protocol lab team. So today we actually have a two section. We want to talk about. uh The first part is we have done some financial analysis uh from a service provider perspective.

A

uh I want to share that with you guys about how you know build your ri model. How do you approach you? Investors and uh you know our our view of um um this whole mining business you know, can hopefully give uh your some guidance regarding you know what are the key elements you need to think about it. What are the risks you want? You need to be aware, and what are the potential area you need to work hard to?

A

You know increase your bottom line, so uh the latter part of there is going to be a little bit more on the architectural um technical side about the scalability and our solution of trying to make your more efficient in terms of operational efficiency and scalability.

A

So the first part we want to talk through the basic assumptions, we're making and then get some basics uh in terms of daily reward. How do we calculate and then we move on to you know if you have investment coming? What's the optimized way of allocating uh for procurement perspectives and how much you're buying for ceiling? How much you need to buy for storage? What's the best strategy there?

A

What is the whole capital expense looks like then we talk about what made of your typical operational expenses and then once we have those we can look start looking at overall return on investment and the profit projection, and the most important part is always the last right we can. We can look at the return on investment scenario and what are the key elements that will impact that rm model left and right and some conclusion or tips that what we should working on strategically.

A

So this stack will really try to understand the question we've been asking ourselves for the last year right, um it's mining. We all believe that something worth to do, but we try to quantify this right. What is my apr return? Investment? uh How about? I just hold the coin instead of uh you know, invest to to mine. If I do mining, what are the things I need to be? You know aware of what are those things? I need to be careful about right.

A

So, like the objective, it is for the fellow service provider to be able to have a better time when they talking to their potential investors, and you know we can build out this community and ecosystem together.

A

Some key key assumptions right because of the equipment depreciation cycle because of the pledge coin and the release schedule to capture all the costs and benefits and return. We extended the period into about 60 months. So you later on you see, a lot of things are calculated over five years of period.

A

That way, we can make sure that all the interests are paid all the pledges coming back and all those type of stuff right. um Another assumptions we're making is all the equipment is, will be fully depreciating three years. I know some miners maybe do four years, but in our environment our servers are running pretty high utilized, so three years is kind of the reasonable expectation later on. You will find out that you know you rather trying to get the new gears right once your equipment is fully depreciated, um one of the most important thing.

A

That is the price right. We all know the price uh it could swing one way or the other, um so the assumption we're taking is you know right now, the price it's low, two digit um last month, when we start this model, it's about 25, so we're making a different scenario right. Consider we were at 240 about a year ago, so I think it's pretty somewhat moderate expectation for it to reach back to 200. In about three years. Of course, we made a simplified assumption.

A

All the price changes are linearly that you know enable us to calculate monthly uh price, but we also want to consider. Maybe the things didn't go very well right. uh Maybe it's a stay in low two-digit, so that's a scenario b. If it only reached two fifties now you know the average is about thirty thirty, seven dollars right over three year period or a better scenario where you know we can reach 500..

A

Personally, I believe, that's probably more likely to happen, but you know you you want to have your bases covered when you talk to your investor and also the worst case scenario, it's even go down to five dollars. In that case, right things really go bad. How was the best way to protect your investment.

A

So when you have some money to spend right to invest what what is the strategy to buy hardware right, um so here is some guiding principle that we come up with. You want to keep your server busy all the time, especially your ceiling capacity, because those are high-end computing assets.

A

So, but you don't want to buy all the disks up front because you don't want them sitting there. You know taking your precious capitals not doing anything, so the storage capacity need to be procured in our model every three months. That means I will have enough ceiling storage ceiling capacity to seal my storage in 90 days, so every quarter I need to buy another bunch of the disc and the storage system, and then my ceiling will be able to seal whatever I buy quarterly in about a 90 days a period.

A

So that means initially my capital will be split between my full ceiling capacity, and my quarterly storage capacity makes sense so far and because of my equipment is going to be fully depreciating three years. That means you know, storage for 180 days. All my storage can be used twice in this three years depreciation cycle and my ceiling can be used for 12 times right because I seal a bunch of storage every quarter, here's how generally how the story being unfold right.

A

So, on the left side, we estimate the building blocks of my mining, my daily return on the right side. It's the pledge required in the middle. It's the capital uh requirement. uh We take you we're assuming you take it. You know usd loans uh with a pretty hefty uh interest rate, uh so we don't want to put any rosy environment right that that's unrealistic.

A

um Also, we have. We uh assume that we also take a loan for the pledge coins. We want to leverage um the services that uh you know, coin lists and the dharmas and those guys put out there to fully take advantage and leverage their plastic loan, and then we can come over all come together as the overall overall return on invest model, and with that we can do some. What? If scenarios and some conclusions.

A

So some of the basics right- those are mathematic uh equations that give us the daily reward uh that will be decaying over time, and these sets of uh formula give us the vesting how much we getting and how much we can get out.

A

How much need to be stayed there for how long and of course the you know, important part is the pledge the pledge requirement there. So, when you look at all that graphically you can see, the daily rewards keep going down. um You know people worry about it. You know, as as time goes by, I'm getting less and less coin.

A

Do I still do this right and this deck will answer that question and how do you hatch that risk? How do you increase right turn this curve back up by you know, store some real data so initially, let's assume you get somewhere between 10 million 11 million, and you have a three bucket units, my money, you need to buy initial down payment for the uh pledge loan that you need to apply for the pledge you need to buy some ceiling capacity. You need to buy some storage capacity.

A

I think I we're talking about this. This slide right. So with initial in investment um somewhere between 10 11 million right, you have a three three bucket of please three bucket of money. You need to spend. You need to spend about three point, one million for the uh down payment of uh pledge, uh so you get enough pledge what we have. A com slide coming up to talk about the terms of those loans, and you also have the rest of the usd dollars right.

A

You need a spend to have a down payment of the equipment purchase and we're also, assuming you you're borrowing, your usd so very important concept. Here I highlighted in the light blue here it's the unit price in terms of capacity right, so right now it's just a number. It's actually have a huge spreadsheet behind it to calculate the unit price uh for in your environment.

A

uh This is a number in our environment, and um so basically uh this is the unit price right we can. We can use that as a building block to scale that up um with that amount of um initial capital, we were able to, as a downpay will be able to buy the first batch of 90 close to 100 petabyte for storage. Right.

A

Remember: we need to be able to seal that storage in 90 days, so we we have about five million dollars as a down payment, so we can buy a little bit over one petabyte of compute that has all the necessary components in there to be able to seal this much of a storage in 90 days.

A

So that's important concept for us to make this model scalable right. If you don't, if you have more money, you simply scale up or less scale that back down right and uh the again the this is a graphic view of how the capital is being expanded to do the hardware procurement. Initially, you need by ceiling power as well as a storage capacity and then moving forward every quarter. You just need to buy additional disks and uh that that, of course, is driven by hardware market right.

A

Exactly what you're buying there's a whole different sense of discussion about the optimized configuration of your sailing? What's the best architecture for your storage? How you could how you derive to this unit cost number, but we're not gonna get down to that level of detail here um on the operational side right operational expense later on, you will find out it's a quite big chunk of changes here. um We also have a very similar concept. The unit cost right. If I have one petabyte of storage. Of course nobody has one petabytes.

A

Most people has more than one petabyte right, maybe 10 20, 50, petabytes of storage, and then you counted all your operational support. Your data center costs your internet equipment maintenance software license subscription. You know your brick and fix your spare parts. um You know the people who keep the lights on your operational staffing.

A

Then you divide it up, um so you get a unit cost of opex right, both from a storage perspective and as well as from sailing perspective, and then, with that unit cost you'll be able to calculate over three years of a full equipment. Depreciation lifecycle, how much money you're going to have to spend. I call that keep the lights on cost right.

A

On top of that, in order for you to mine, you will need spend some money initially to buy the fail to use as a guest fee and later on, once you have uh reward, the reward will use to cover the gas. When we have all this equipment in place right, we're buying, we keep buying every quarter and then, each month, how much I can sell sale and how much I need from pledge over about six uh 60 months. This is the number of coins I need as a pledge.

A

um So this is this is what I would use to talking to uh people like a derma and anchorage and right. This is my projection of how many loans I need to take.

A

um So, of course, we need to talk about the financing cost right. um People willing to uh loan you money so for the usd um the assumption we're taking is annual interest is 15. That's pretty high. um We needed 20 of down payment, uh so initially we're we'll have seven million dollars as a down payment, but uh since we're buying additional storage right over three years period, this is down payment we put down, and this is total amount. We are borrowing- and this is the interest were paid and interest is part of the cash flow.

A

Similar on the fail side right, we need a. We need a loan for the pledge, uh because of the 180 days of nature, of that uh we will need that for about year and a half uh same interest rate same down payment. um You know this is the number of fulfilled. We need to pay um as a down payment, because initially you do not have fail. You have to use usd to to buy fail right, uh the amount of total you borrow and the interest you pay on those fail.

A

Why do we trying to take a failed pledge loan? Because that's that's going to hatch the risk about the price change you take the loan in fail. You return. The loan fail right, uh the price you are immune from the price change to a large degree.

A

So now, with all the money right um so over three years, you, if you add them all up this- is how much you spend total um hardware cost 30 on operation cost and 17 on the interest.

A

So if we apply that kind of numbers over three year period, not three or five years, rather, this is roughly the intensity of the capital being spended months by months. Right, as you can see the capex uh initially, then you drop down. Then you start.

A

I I take the financial interest out because that's a you know decent amount of number to to to have its own uh column right, but you notice that the opex are increasing, because your footprint is increasing right um and then it's a stay at a certain rate and that's dropped down because your equipment started to be decommissioned.

A

Note this number right: you total spending is about 187 million dollars for for this model for this type of capacity, and you get close to 600 petabyte of the storage on the the previous slide is on the usd dollar right, um but on the fail side right every month, you have unless first a couple of months right once you start have a sealed capacity. You start generating coins and your coins keep getting up right and you.

A

You also need to return your principal payback uh of your pledge loans and your pledges start to return right once you add them all up. This is overall accumulative fail. You are getting right throughout the entire five-year period. You are getting.

A

6 million 6.2 million fails and consider everything you spend remember that 187 million dollars you spent your average cost of fail is 33. Keep that number in mind. This is very important because that is a break-even point. That means I spend all this money. I got all this number for fail.

A

My average cost of obtaining a fail is about 33 dollars. So remember that assumption we made on the fair price if the price is above, that you're profitable it's below that you're, not right so we'll later on talk about those scenarios.

A

So this is a monthly profit. Now we apply that scenario. We talked about earlier right. uh The price stay at the 50, not stay at 50, but it slowly increased to 50 or moderate increase to 200, or you know, everyone look to see grab uh to 500 or the worst case job down to five.

A

So this is the monthly profit.

A

An assumption we're making is any extra fail you gained at that month after you pay your gas after you pay your interest. If you still have something left over, you sell that at the market price, so just so that we can compare the usd for the um for ri right.

A

So important question we want to answer here: is you know for potential investors who look good for this whole market, the fair coin ipfs right? They want to invest, but the question they asked is: why do I want to invest a minor versus? I just buy a bunch of a failed coin and hold it right. uh This slide answered that question. If you just hold it assuming remember the go back, a few slide.

A

Remember the capital spending schedule we talked about. If you, instead of spending that capital to buy equipment to mine, you use the same spending intensity same schedule to buy coin at the same assumption, we're making right if you're, simply holding the coin.

A

Your average cost of holding the coin is, of course, increased because the coin price keeps increasing right. But if you mine, you're, pretty much immune from the price fluctuation a little bit, because you still need to buy initial down payment, you still need to buy gas whatever right, but you pretty much stay at a 33 dollar range.

A

This is important. That means my cost per fill stay at 33. If I believe next three years, the average cost is above that anything above that is my profit.

A

But if I keep buying coin I'm vulnerable because I don't control at the price. If prices go high, my costs go high.

A

So another way to look at that is.

A

Now we everything so far, we talk about it. It's no real data right, if we add real data to it, fail plus zero percent real data, five percent real data about one third of the data or majority of real data right so with real data, because your return is 10 times more, so even in the worst case scenario, in the worst case scenario, where everything else is losing money, but if you have enough real data that will protect you from losing money.

A

Of course, in the you know, all the other scenario, real data will greatly boost your return, some of conclusions, um so even in the worst case scenario, where you know 25 dollars per fail, drop down to five dollars three years from now right, real data is your saver um and, of course, if you just holding the fail, you know that in the in that worst case scenario, you don't have much hold on to right.

A

But if you have invest large amount of infrastructure, you still have a bunch of a server that has residual value and all that right. So it's actually better from the risk mitigation perspective and uh the break-even point right in this model is about 33 dollars and, like you said earl like you, you see earlier right.

A

That is pretty much immune from the fluctuation of the fail price. So, no matter fair price go high. Your cost is still about 33 dollars, because your cost of mining roughly stay the same right and the most important thing is get on to the failed plus, because that's going to significantly boost your ri.

A

And um one more point I want to make is, if you remember that pie chart 30 of that, because you don't have much control over the loan uh terms and the interest rate you don't have much control over the capital market versus you know the equipment market, but you do have control over how you run your ship right. How tightly your operational efficiency is, uh which kind of lead to my next presentations about? You know the software defined infrastructure. But the point is you: your opex is under your control right.

A

If you run your entire stacks very efficiently. That directly goes to your bottom line.

A