Tony Nash joins Daniel Lacalle in a discussion on the rise of the machines in a form of AI and machine learning and how Complete Intelligence helps clients automate budgeting with better accuracy using newer technologies like now casts. How GDP predictions are actually very erroneous yet nobody gets fired? And how about China’s GDP as well, and why it’s different from other economies? All these and so much more in markets in this fun discussion.
The video above is published by Daniel Lacalle – In English.
DL: Hello everyone and welcome to this podcast. It is a great pleasure to have somebody that you should actually follow in social media on Twitter, Tony Nash. He is somebody that you definitely need to need to look for because it has very very interesting ideas. Tony, how are you?
TN: Great, thanks Daniel. Thanks so much for having me today.
DL: It’s a tremendous pleasure as I said I was very much looking forward to to have a chat with you. Please introduce a little bit yourself. A little bit to our audience and let us know what is it that you do.
TN: Sure, thanks Daniel. My name is Tony Nash. I live in Houston, Texas. I’ve spent actually most of my life outside of the U.S. I spent most of my 20s in Europe, North Europe, the UK, Southern Europe and from my 30s to almost the end of my 40s I was in Asia. And so you know being in the U.S., Europe and Asia has really given me personally an interesting view on things like trade economics markets and so on and so forth.
During that time I was the global head of research for the economist out of London, I was based in Singapore at the time. Led the global research business. I moved from there to lead Asia consulting for a firm called IHS Markit which is owned by S&P now.
And after that I started my current firm Complete Intelligence which is a machine learning platform. We do global markets currencies, commodities, equity indices, economic concepts. We also do corporate revenue and expense forecasting so we’ll automate budgeting for large multinational firms.
DL: Wow! amazing. Truly amazing. You probably have a very interesting viewpoint on something that a lot of the people that follow us have probably diverging views. Know the situation about the impact of algorithms in the market the impact of high frequency trading and machines in markets.
We had a chat a few months ago with a professor at the London School of Economics that he used to invite me to his year-end lectures to to give a master class. And he mentioned that he was extremely concerned about the almost the rise of the machines. What is your view on this?
TN: I think so an Algo is not an Algo, right? I mean, I think a lot of the firms that are using Algo’s to trade are using extremely short-term algorithmic trading say horizons. Okay? So they’re looking at very short-term momentum and so on and so forth. And that stuff has been around for 10 plus years, it continues to improve. That’s not at all what we do we do monthly interval forecasts, Okay?
Now, when you talk to say an economist they’re looking at traditional say univariate and multivariate statistical approaches, which are kind of long-term trendy stuff. It’s not necessarily exclusively regression, it gets more sophisticated than that.
When we talk to people about machine learning, they assume we’re using exclusively those kind of algorithms. It’s not the case. There’s a mix we run what’s called an ensemble approach. We have some very short-term approaches. We have some longer-term traditional say econometric approaches. And then we use a configuration of which approach works best for that asset or that revenue line in a company or that cost line or whatever for that time.
So we don’t have let’s say, a fixed Algo for gold, Okay? Our algorithm for the gold price is continually changing based upon what’s happening in the market. Markets are not static, right? Trade flows economics, you know, money flows whatever they’re not static. So we’re taking all of that context data in. We’re using all of that to understand what’s happening in currencies, commodities and so on, as well as how that’s impacting company sales. Down to say the department or sub department level.
So what we can do with machine learning now. And this is you know when you mentioned should we fear the rise of the machines. We have a very large client right now who has hundreds of people involved in their budgeting process and it takes them three to four months to do their budgeting process. We’ve automated that process it now takes them 72 hours to run their annual budgeting process, okay? So it was millions of dollars of time and resources and that sort of thing. We’ve taken them now to do a continuous budgeting process to where we churn it out every month. So the CFO, the Head of FP&A and the rest of the say business leadership, see a refresh forecast every month.
Here’s the difference with what we do, compared to what a lot of traditional forecasters and machine learning people do, we track our error, okay? So we will as of next month have our error rates for everything we forecast on our platform. You want to know the error for our gold price forecast, it’ll be on there. You’ll know the error for our Corn, Crude, you know, JPY whatever, it’s on there. So many of our clients use our data for their kind of medium term trades so they have to know how to hedge that trade, right? And so if we have our one, three month error rates on there, something like that it really helps them understand the risk for the time horizon around which they’re trading. And so we do the same for enterprises. We let them know down to a very detailed level to error rates in our forecast because they’re taking the risk on what’s happening, right? So we want them to know the error associated with what they’re doing with what we’re doing.
So coming out of my past at the economist and and IHS and so on and so forth. I don’t know of anybody else who is being transparent enough to disclose their error rates to the public on a regular basis. So my hope is that the bigger guys take a cue from what we’re doing. That customers demand it from what we’re doing. And demand that the larger firms disclose their error rates because I think what the people who use information services will find is that the error rates for the large firms are pretty terrible. We know that they’re three to seven times our error rates in many cases but we can’t talk about that.
DL: But it’s an important thing. What you’ve just mentioned is an important thing because one of the things that is repeated over and over in social media and amongst the people that follow us is well, all these predictions from the IMF, from the different international bodies not to the IMF. Actually the IMF is probably one of the one that makes smaller mistakes but all of these predictions end up being so aggressively revised and that it’s very difficult for people to trust those, particularly the predictions.
TN: Right. That’s right.
DL: And one of the things that, for example when we do now casts in our firm or when with your clients. That’s one of the things that very few people talk about, is the margin of error is what has been the mistake that we have made in the in that previous prediction. And what have we done to correct it because one might probably you may want to expand on this. Why do you think that the models that are driving these now cast predictions from investment banks in some cases from international bodies and others? Are very rarely revised to improve the prediction and the predictability of the of the figures and the data that is being used in the model.
TN: It’s because the forecasters are not accountable to the traders, okay? One of the things I love about traders is they are accountable every single day for their PNO.
DL: Yeah, right.
TN: Every single day, every minute of every day they’re accountable for their PNO. Forecasters are not accountable to a PNO so they put together some really interesting sophisticated model that may not actually work in the real world, right? And you look at the forward curves or something like that, I mean all that stuff is great but that’s not a forecast, okay? So I love traders. I love talking to traders because they are accountable every single day. They make public mistakes. And again this is part of what I love about social media is traders will put their hypothesis out there and if they’re wrong people will somewhat respectfully make fun of them, okay?
DL: Not necessarily respectfully but they will.
TN: In some cases different but this is great and you know what economists and industry forecasts, commodity forecasters these guys have to be accountable as well. I would love it if traders would put forecasters up to the same level of criticism that they do other traders but they don’t.
DL: Don’t you find it interesting? I mean one of the things that I find more intellectually dishonest sometimes is to hear some of the forecasters say, well we’ve only made a downgrade of one point of one percentage point of GDP only.
TN: Only, right. It’s okay.
DL: So that is that we’ve grown accustomed to this idea that you start the year with a prediction of say, I don’t know three percent growth, which goes down to below two. And that doesn’t get anybody fired, it’s sort of like pretty much average but I think it’s very important because one of the things. And I want to gather your thoughts about this. One of the things that we get from this is that there is absolutely no analysis of the impact of stimulus packages for example, when you have somebody is announcing a trillion dollar stimulus package that’s going to generate one percent increase in trendline GDP growth it doesn’t. And everybody forgets about it but the trillion dollars are gone. What is your thoughts on this?
TN: Well, I think those are related in as much as… let’s say somebody downgraded GDP by one percent. What they’re not accounting for, What I think they’re not accounting for is let’s say the economic impact kind of multiplier. And I say that in quotes for that government spending, right? So in the old days you would have a government spending of say you know 500 billion dollars and let’s say that was on infrastructure. Traditionally you have a 1.6 multiplier for infrastructure spend so over the next say five years that seeps into the economy in a 1.6 times outs. So you get a double bang right you get the government spending say one-to-one impact on the economy. Then you get a point six times that in other industries but what’s actually happened.
And Michael Nicoletos does some really good analysis on this for China, for example. He says that for every unit of say debt that’s taken out in China, which is government debt. It takes eight something like eight units of debt to create one unit of GDP. So in China for example you don’t have an economic multiplier you have an economic divisor, right?
TN: So the more the Chinese government spends actually the less GDP growth which is weird, right? But it tells me that China is an economy that is begging for a market. A real market, okay? Rather than kind of central planning and you and Europe. I’m sure you’re very familiar with the Soviet Union. I studied a lot of that in my undergrad days very familiar with the impact of central planning. China there’s this illusion that there is no central planning in China but we’re seeing with the kind of blow-ups in the financial sector that there is actually central planning in China.
And if you look at the steel sector you look at commodity consumption, these sorts of things it’s a big factor of china still, right? So but it’s incredibly inefficient spending. It’s an incredibly inefficient way and again it’s a market that is begging for an open economy because they could really grow if they were open but they’re not. They have a captive currency they have central planning and so on and so forth.
Now I know some of the people watching, you’re going to say you’ve never been to China, you don’t understand. Actually I have spent a lot of time in China, okay? I actually advise China’s Economic Planners for about a year and a half, almost two years on the belt and road initiative. So I’ve been inside the bureaucracy not at the high levels where they throw nice dinners. I’ve been in the offices of middle managers for a long time within the Chinese Central Government so I understand how it works and I understand the impact on the economy.
DL: Don’t you think it’s interesting though that despite the evidence of what you just mentioned. And how brutal it has been because it’s multiplied by 10. How many units of debt are required to generate one unit of GDP in a little bit more than a decade? Don’t you find it frustrating to read and hear that what for example the United States needs is some sort of central planning like the Chinese one. And that in fact the the developed economies would be much better off if they had the type of intervention from from the government that China has?
TN: Sure, well it’s it’s kind of the fair complete that central bankers bring to the table. I have a solution. We need to use this solution to bring fill in the blank on desired outcome, okay? And so when central bankers come to the table they have there’s an inevitability to the solution that they’re going to bring. And the more we rely on central bankers the more we rely on centralized planning. And so I’ve had so many questions over the last several years, should the us put forward a program like China’s belt and road program, okay?
We know the US, Europe, the G20 nobody needs that, okay? Why? Because Europe has an open market and great companies that build great infrastructure. The US has an open market and although European infrastructure companies are better. The US has some pretty good companies that build infrastructure in an open market. So why do we need a belt and road program? Why do we need central planning around that? And we can go into a lot of detail about what’s wrong with the belton road and why it’s not real, okay? But that type of central planning typically comes with money as the as kind of the bait to get people to move things. And so we’re already doing that with the FED and we’re already doing that with treasure with money from the treasury, right?
And if you look at Europe you’re doing it with the ECB. You’re doing it with money from finance ministries. The next question is, does the government start actually taking over industries again? And you know maybe not and effectively in some ways they kind of are in some cases. And the real question is what are the results and I would argue the results are not a multiplier result they are a divisor result.
DL: Absolutely. Absolutely it is we saw it for example. I think it’s, I mean painfully evident in the junk plan in Europe or the growth and jobs plan of 2009 that destroyed four and a half million jobs. It’s not easy to to achieve this.
TN: You have to try to do that.
DL: You have to really really try it, really try.
I think that you mentioned a very important factor which is that central banking brings central planning because central banks present a program of monetary easing of monetary policy. And they say well we don’t do fiscal policy but they’re basically telling you what fiscal policy has to be implemented to the point that their excuse for the lack of results of monetary policy tends to be that the that the transmission mechanism of monetary policy is not working as it should. Therefore because the demand for credit is not as much as the supply of money that have invented. They say, well how do we fill in the blank? Oh it has to be government spending. It has to be for planning. It has to be so-called infrastructure spending from government.
You just mentioned a very important point there is absolutely no problem to invest in infrastructure. There’s never been more demand for a good quality infrastructure projects from private equity, from businesses. But I come back to the point of of central banks and a little bit about your view. How does prolonging easing measures and maintaining extremely low rates affect these trends in growth and in these trends in in productivity?
TN: Well, okay, so what you brought up about central banks and the government as the transmission mechanism is really important. So low interest rates Zerp and Nerp really bring about an environment where central banks have forced private sector banks to fail as the transmission mechanism. Central banks make money on holding money overnight, that’s it. They’re not making money necessarily or they’re not doing it to successfully to impact economies. They’re not successfully lending out loans because they say it’s less risky buying bonds. It’s less risky having our money sit with the Fed. It’s less risky to do this stuff than it is to loan out money. Of course it’s less risky, right? That’s goes without saying.
So you know I think where we need to go with that is getting central banks out of that cycle is going to hurt. We cannot it… cannot hurt, well I would say baby boomers in the West and and in Northeast Asia which has a huge baby boomer cohort. Until those guys are retired and until their incomes are set central banks cannot take their foot off the gas because at least in the west those folks are voters. And if you take away from the income of that large cohort of voters then you’ll have, I guess I think from their perspective you’ll have chaos for years.
So you know we need to wait until something happens with baby boomers. You tell central banks and finance ministries or treasuries will kind of get religion and what will happen is behind baby boomers is a small cohort generally, okay? So it’s that small cohort who will suffer. It’s not Baby Boomers who will suffer. It’s that small cohort who will suffer. It’s the wealth of that next generation that Gen x that will suffer when central banks and finance ministries get religion.
So we’re probably looking at ten more years five more years of this and then you’ll see kind of… you remember what a rousing success Jeff Sax’s shock therapy was, right?
TN: So of course it wasn’t and it’s you know but it’s gonna hurt and it’s gonna hurt in developed countries in a way that it hasn’t hurt for a long time. So that kind of brings to the discussion things like soundness of the dollar, status of the Euro that sort of thing. I think there are a lot of people out there who have this thesis. I think they’re a little early on it.
DL: Yeah, I agree.
TN: So economists you know these insurance people see it from a macro perspective but often they come to the conclusion too early. So I think it’s a generational type of change that’ll happen and then we start to see if the US wants the dollar to remain preeminent. They’re going to have to get religion at the central bank level. They’re going to have to get religion at the fiscal level and really start ratcheting down some of the kind of free spending disciplines they’ve had in the past.
DL: Yeah, it’s almost inevitable that you’re in a society that is aging. The net prison value of bad decisions for the future is too positive for the voters that are right now with the middle age, in a certain uh bracket of of age. Me, I tried the other day my students I see you more as the guys that are going to pay my pension than my students. So yeah…
TN: But it’s you and me who will be in that age bracket who will pay for it. It’s the people who are 60 plus right now who will not pay for it. So they’ll go through their lives as they have with governments catering to their every need, where it’s our age that will end up paying for it. So people our age we need to have hard assets.
TN: You know when the time comes we have to have hard assets because it’s going to be…
DL: That is one of one of the mistakes that a lot of the people that follow us around. They they feel that so many of the valuations are so elevated that maybe it’s a good time to cash in and simply get rid of hard assets, I say absolutely the opposite because you’ve mentioned a very important thing which is this religious aspect that we’ve that we’ve gotten into. And I for just for clarity would you care to explain for people what that means because…
TN: I say get religion? I mean to become disciplined.
DL: I know like you because that is an important thing.
TN: Yes, sorry I mean if anybody but to become disciplined about the financial environment and about the monetary environment.
DL: Absolutely because one of the things that people tend to believe when you talk about religion and the the state planners religion and and central bank’s religion is actually the opposite. So I wanted to write for you to very make it very clear. That what you’re talking about is discipline you’re not talking about the idea of going full-blown MMT and that kind of thing.
TN: No. I think if there is if there is kind of an MMT period, I think it’s a I don’t think it’s an extended period. I think it’s an experiment that a couple of countries undertake. I think it’s problematic for them. And I think they try to find a way to come back but…
DL: How do you come back from that because one of the problems that I find when people bring the idea of well, why not try. I always, I’m very aware and very concerned about that thought process because you know I’ve been very involved in analyzing and in helping businesses in Argentina, in Hawaii, in Brazil and it’s very difficult to come back. I had a discussion yesterday with the ex-minister of economy of Uruguay and Ignacio was telling me we started with a 133 percent inflation. And we were successful in bringing it down to 40 and that was nine years.
TN: Right. So, yeah I get how do you come back from it look at Argentina. look at Zimbabwe. I think of course they’re not the Fed. They’re not you know the EU but they are very interesting experiments when people said we’re going to get unhinged with our spending. And we’re going to completely disregard fundamentals. Which I would say I would argue we are on some level disregarding fundamentals today but it’s completely you know divorced from reality. And if you take a large economy like the US and go MMT it would take a very long time to come back.
TN: So let’s let’s look at a place like China, okay? So has China gone MMT? Actually, not really but their bank lending is has grown five times faster than the US, okay? So these guys are not lending on anything near fundamentals. Sorry when I say five times faster what I mean is this it grew five times larger than the bank lending in the US, okay? So China is a smaller economy and banks have balance sheets that are five times larger than banks in the US. And that is that should be distressing followers.
DL: Everybody say that the example of China doesn’t work because more debt because it’s growing faster what you’ve just said is absolutely critical for for some of our followers.
TN: Right, the other part about China is they don’t have a convertible currency. So they can do whatever they want to control their currency value while they grow their bank balance sheets. And it’s just wonderland, it’s not reality so if that were to happen there are guys out there like Mike Green and others who look at a severe devaluation of CNY. And I think that’s more likely than not.
DL: Yeah, obviously as well. I think that the the Chinese government is trying to postpone as much as it can the devaluation of the currency based on a view that the imbalances of the economy can be sort of managed through central planning but what ends up happening is that you’re basically just postponing the inevitable. And getting a situation in which the actual devaluation when it happens is much larger. It reminds me very much. I come back to the point of Argentina with the fake peg of the peso to the dollar that prolonging it created a devastation from which they have not returned yet.
TN: Right. And if you look at China right now they need commodities desperately, okay? Metals, they need energy desperately and so on and so forth. So they’ve known this for months. So they’ve had CNY at about six three, six four to the dollar which is very strong. And it was trading a year ago around seven or something like that. So they’ve appreciated it dramatically and the longer they keep it at this level. The more difficult it’s going to be on the other side. And they know it these are not stupid people but they understand that that buying commodities is more important for their economy today because if people in China are cold this winter and they don’t have enough nat gas and coal then it’s going to be a very difficult time in the spring for the government.
DL: And when you and coming back to that point there’s a double-edged sword. On the one side you have a currency that is out to free sheet are artificially appreciated. On the other side you also have price controls because coal prices are limited by the government. And therefore you’re creating on the one hand a very big monetary hole and on the other hand a very big financial hole in the companies that are selling at a loss.
TN: That’s true but I would say one slight adjustment to that things like electricity prices are controls. When power generators buy coal, they buy that in a spot market, okay? So coal prices have been rising where electricity prices are highly regulated by the government this is why we’ve seen blackouts and brownouts and power outages in China. And why it’s impacted their manufacturing base because they’re buying coal in a spot market and then they’re having to sell it at a much lower price in the retail market.
And so again this is the problem with central planning this is the problem with kind of partial liberalization of markets. You liberalize the coal price but you keep the electricity price regulated and if you don’t have the central government supporting those power plants they just blow up all over the place. And we’ve seen the power generators in the UK go bankrupt. We saw some here in Texas go bankrupt a couple years ago because of disparities like that and those power generators in the UK going bankrupt that’s the market working, right? So we need to see that in China as well.
DL: Yeah, it’s a very very fascinating conversation because on the other hand for example in Europe right now with the energy shortage we’re seeing that a few countries Spain, France, etc. are actually trying to convince the European Union, the European Commission to try to get into a sort of intervened market price in the in the generation business. Which would be just like you’ve mentioned an absolute atrocity very very dangerous.
TN: This creates a huge liability for the government.
DL: It creates a massive liability for the government. This is a key point that people fail to understand the debate in the European union is that, oh it’s a great idea because France has this massive utility company that is public. And therefore there’s no risk it had to be bailed out twice by the taxpayers. People tend to forget that you’re paying for that.
TN: But again this is what’s that block of voters who doesn’t really care about the impact 10 or 20 years down the road. That’s the problem. There’s a huge block of voters who don’t really care what the cost is because the government’s going to borrow money long-term debt. And it’s going to be paid back in 10 or 20 years and the biggest beneficiaries of this and the people on fixed incomes they actually don’t care what the cost is.
DL: Yeah, yeah exactly, exactly. There’s this fantastic perverse incentive to pass the bill to the next generation. And that obviously is where we are right now. Coming back to the point of the infrastructure plans and the belt and road plan. What in your view are the the lessons that we must have learned or that we should be learning from the Belgian road initiative?
TN: So here’s a problem with the Belton road and I had a very candid discussion with a senior official within China’s NDRC in probably 2015 which was early on, okay? And this person told me the following they said the Belgian road was designed to be a debt financed plan. What’s happening now, and again this was six or seven years ago, very early on in the in the belts and road dates. They said the beneficiary countries are pushing back and forcing us to take equity in this infrastructure, okay?
Now why does that matter well the initial build out of infrastructure is about five percent of the lifetime cost of that asset, okay? So if you’re if China is only involved in the initial build out they’re taking their five percent, it’s a loan and they get out. If they’re equity holders in that let’s say they’re 49 equity holders in an Indonesian high-speed rail then they become accountable for part of that build-out. And then they have to maintain the other 95 of the cost for the next 30 to 50 years. So they thought they were going to be one and done in and out. We do this infrastructure we get out they owe us money and it’s really clean what’s happened is they’ve had to get involved in the equity of those assets.
And so I’ve since had some uh government officials from say Africa ask me what do we do with the Belton road with china? Very simple answer force them to convert the debt to equity, okay? They become long-term involved on a long-term basis. They become involved in those assets and then they’re have a different level of interest in them in the quality maintenance and everything else but they’re also on the long-term basis accountable for the costs.
So they don’t just build a pretty airport that and I’m not saying this necessarily happens but they don’t just build a pretty airport that falls apart in five years, okay? They then have to think about the long-term impacts and long-term maintenance costs of that airport, right? And so but you know the original design of the Belton road was debt financing. Mobilizing workers and so on and so forth what it’s become is a mix of debt and equity financing. And that’s not what the Chinese government has wanted.
So I’ve been telling people for three or four years the Belton road is dead, okay? And people push back me and say no it’s not, you know think tank people or whatever. But they don’t understand the fundamental fact of how the Belton road was designed it was designed as a one-and-done debt financed infrastructure build out it’s become a long-term investment all around the world. So it’s a different program. It’s failed, okay?
They’re not going to make the money they thought yes they’ll keep some workers busy but they’re not going to make the money they thought. All of those assets, almost all those assets are financed in US dollars, okay? So they’re not getting their currency out. It’s not becoming an international unit like they had hoped. They’re it’s not they’re not clean transactions and so on and so forth. So this is what’s happened with the Belgian road. So the lesson learned is they should have planned better. And they should have had a better answer to you become an equity owner. And uh
I think you know if any western governments want to have kind of a belt and road type of initiative. They’re going to have to contend with the demand from some of these countries that they become equity owners. And I think that’s a bad idea for western governments to be equity owners in infrastructure assets so you know this is this is the problem.
Japanese have taken a little bit different because of where the Yen is and because of where interest rates are in Japan. Japanese have basically had kind of zero interest or close to zero interest on the infrastructure they’ve built out. And so they haven’t gone after it as aggressively as China has. They’ve had a much cleaner um structure to those agreements. And so they’ve been, I think pretty successful in staying out of the equity game and staying more focused on the debt financing for their infrastructure initiatives.
DL: Oh, absolutely big lesson, big lesson there because the we see now that the vast majority of those projects are impossible to the debt is impossible to be repaid. There’s about 600 billion dollars of unpayable debt out there. And we also have the example from from the internationalization of the French, Spanish, Italian companies into Latin America that they fell into the same trap. They started with a with a debt-financed infrastructure build type of clean slate program that ended up owning equity. And in some cases with nationalizations hopefully that will not…
TN: And watch for debt to equity conversions in these things. It’s good. There’s going to be huge pressure because the Chinese say the exit bank the CDB. A lot of these organizations are going to be forced to convert that debt to equity and then unload it on operating companies in China. They’re not going to want to do it but we’re going to start to see more and more pressure there over the next couple of years.
DL: Great! Well I’m absolutely convinced that will happen. Tony, we’ve run out of time so it’s been an incredible conversation lots of things that are very very interesting for our followers. We will give all the details to follow you and to get more information about your company in the details of the of the video. And thank you so much for your time. I hope that that we will be able to talk again in a not too distant future.
TN: Thank you Daniel. Anytime. Thank you so much.