首页  >>  来自播客: Leaders in Lending 更新   反馈

The Upstart Macro Index (UMI)

发布时间 2023-03-29 18:00:02    来源

摘要

In 2021, loan repayments were at a historic high, but by mid-2022 the tables had turned. What happened?To get to the bottom of this alarming shift, Upstart created the Upstart Macro Index (UMI), a precise, quantitative method for separating the micro and macro effects that should be influencing underwriting decisions today.In this episode, we speak with Paul Gu, Co-Founder and Head of Product at Upstart, about the index and the insights it can offer lenders.Join us as Jeff and Paul discuss:The concept of the Upstart Macro Index (UMI)Micro versus macro prediction levelsHow to leverage UMI in the context of lending partnershipsWant to learn more about the Upstart Macro Index (UMI)? Check out the UMI site mentioned in this episode.

GPT-4正在为你翻译摘要中......

中英文字稿  

You're listening to leaders in lending from Upstart, a podcast dedicated to helping consumer lenders grow their programs and improve their product offerings.
你正在收听来自Upstart的贷款领袖，这是一个致力于帮助消费者贷款人士增加他们的计划和改进产品提供的播客节目。

Each week, hear decision makers in the finance industry offer insights into the future of the lending industry. Best practices around digital transformation. And more, let's get into the show.
每周，听听金融业的决策者们对借贷行业未来的见解。最佳数字化转型实践等等，让我们开始节目吧。

Welcome to Leaders in Lending. I'm your host, Jeff Keltner. This week's episode features my conversation with Paul Gooe, the co-founder and the head of product and data at Upstart.
欢迎来到借贷领导者。我是你们的主持人Jeff Keltner。本周的节目中，我将和Upstart的联合创始人、产品和数据主管Paul Gooe进行对话。

Actually, this is our second conversation with Paul. And his first episode is actually still our most listened to episode of all time. So we're excited to bring Paul back.
实际上，这是我们第二次与保罗进行的对话。而他的第一篇文章实际上仍然是我们最受欢迎的节目。因此，我们很高兴再次邀请保罗回来。

It was a really interesting conversation delving into some recent innovations. Upstart is made around measuring the impact of macroeconomic conditions on loan performance.
这次谈话真是有趣，深入探讨了一些最近的创新。Upstart是围绕着衡量宏观经济条件对贷款表现的影响而建立的。

Really separating out individual performance of loans from the macroeconomies impact on those in a very quantifiable way, which I thought was a fascinating topic. It's a really interesting area of research.
我认为，从宏观经济的影响中非常明确地分离出贷款个体表现的方式是一项非常引人入胜的研究课题。这是一个非常有趣的研究领域。

So we kind of dive into how it works, why, how that can be utilized and what Upstart is working on in the future towards us.
所以我们会深入探讨它如何运作，为什么，如何利用以及Upstart未来对我们的发展所做的工作。

It's a really interesting topic, one, I obviously highly relevant in the current environment. And so I appreciated having a chance to dive deep with it with Paul.
这个话题非常有意思，显然在当前的环境下非常相关。因此，我很感激有机会跟保罗一起深入探讨它。

I will say there's one correction to make up front, which is that Paul mentions what he called the Upstart macro forecast or a UMF. And the correct term is Upstart macro factor, still UMF. So just a quick correction.
我要说的是，前面有一个需要更正的地方，就是Paul提到了他所谓的Upstart宏观预测或UMF。而正确的术语应该是Upstart宏观因素，仍然是UMF。所以只是一个快速的更正。

And without further ado, please enjoy this conversation with Paul Gooe.
不多说了，请享受这次与保罗·古的交谈。

Paul, welcome back to the podcast. Thanks for joining us again.
保罗，欢迎回到播客节目。感谢您再次加入我们。

Oh, I'm happy to be back. It's grown quite a bit since I was last here.
哦，我很高兴回来了。自从我上次来这里以后，它已经变化了很多。

It is. It is one of the most popular episodes, the original that we've ever had. So apparently already it's like technical discussions. We talked about AI and lending and a more technical way last time. So we're going to try and give them a little bit more technical stuff.
这是其中一个最受欢迎的原始剧集之一。显然我们已经开始涉及技术讨论。上次我们在谈论人工智能和贷款等技术方面。所以我们要试着给观众提供更多的技术内容。

Great. Let's do it.
太好了，我们开始做吧。

Well, the topic of discussion today that I really wanted to dive into is the topic of what we call the Upstart macro index or UMF. Because that's been, I know we've talked about it a bunch and earnings.
今天我们想要深入探讨的话题是我们所说的Upstart宏观指数或UMF。我知道我们已经谈了很多次关于收益方面的问题，但这个话题非常重要。

We talked about it publicly. We've talked about it with our partners. But we haven't disclosed a lot about it or gone in depth. And so I know some people have gotten one-on-one introductions or explanations. But I would have spent the day kind of diving into it.
我们已经公开讨论过这个话题，并与我们的合作伙伴进行了沟通。但我们没有透露太多细节或深入探讨。有些人已经得到了一对一的介绍或解释。但我会花一整天深入了解它。

So maybe the best way to just start is like, what is UMI?
也许最好的方法就是先从什么是UMI开始了解吧？

Yeah. UMI stands for Upstart macro index. And before I describe what it is, I might take a step back and just describe how we got here and why we started working on this thing that became known as the UMI.
好的。UMI 是 Upstart 宏指数的缩写。在我描述它是什么之前，我可能需要退后一步，简要描述一下我们是如何到达这里的以及为什么我们开始着手处理这个被称为 UMI 的项目。

If you take a step back and look at where we were in, call it, mid-2021. At that time, very loosely speaking, almost everybody was doing a good job of repaying their looks. Obviously not everybody. There were always defaults. But repayment rates were extremely high compared to historical norms.
如果你退后一步，看看我们在2021年中期时的状态。那时，非常笼统地说，几乎所有人都在很好地偿还他们的贷款。当然并非所有人都是如此。总会有拖欠情况。但与历史水平相比，还款率非常高。

And it had looked like at the time everybody was a genius in lending. You could have lent to pretty much any type of consumer and probably you are outperforming your expectations in terms of repayment rates and therefore rates of return on debt capital deployed.
当时似乎每个人都是放贷方面的天才。你可以向几乎任何类型的消费者借款，可能会超出你的预期，得到更高的还款利率和债务资本回报率。

Then you fast forward another year to mid-2022 and the story had flipped. Instead of every type of borrower outperforming, now you had almost every type of borrower underperforming.
然后你快进到2022年中期，故事发生了反转。与每种类型的借款人表现良好相反，现在几乎每种类型的借款人都表现不佳。

And the change from mid-2021 to mid-2022 was dramatic. And you could see that in a variety of metrics. But it wasn't totally clear how you would precisely tease that out and quantify what had happened in the quote unquote macro environment from mid-21 to mid-22.
从2021年中期到2022年中期的转变是戏剧性的，从各种指标上可以看出。但是我们还不完全清楚如何准确地分析和量化从21年中期到22年中期所发生的所谓的宏观环境的变化。

We started to have conversations with a lot of our lending partners of the forum. Well, what's going on? All I see is that delinquency rates have gone through the roof. I don't know if it's because delph start decided that has delph start to model has gone bad. Delph start to sort of open the floodgates to low-quality borrowers. What servicing degraded. What in the world is going on?
我们开始和很多论坛的贷方合作伙伴进行交流。嗯，怎么回事？我看到的只是拖欠率飙升。我不知道是不是因为Delph开始决定使用的Delph开始模型出了问题。Delph开始向质量较低的借款人敞开了大门，造成了服务质量的下滑。究竟发生了什么？

And when you see a very dramatic change 2X3X, they weren't talking large multiple in a short period of time. People naturally get worried and they want to know. And of course, we wanted to know too.
当你看到极为剧烈的变化2X3X时，他们并不是在短时间内谈论大幅度的提高。人们自然会变得担心并想要了解。当然，我们也想知道。

And we had a pretty good sense from the indirect metrics that we had, of course, kept our eye on. But it wasn't precise in the way that we would like to have had it. And so we said, well, we really want a precise quantitative way to tease apart what we call the macro and the micro effects of underwriting.
我们从我们关注的间接指标中得到了相当不错的感觉，但它不像我们想要的那样精确。因此，我们说，我们真的想要一种精确的定量方式来分开我们所谓的承保的宏观和微观影响。

And more precisely what I mean by that is if you think about what an underwriting model does, the one hand it has to separate risk. Meaning it has to say this particular applicant is more likely to default than this other one and they're more likely to default by 2X.
更准确地说，我的意思是，如果您考虑一下承销模型的作用，一方面，它必须区分风险。这意味着它必须说这个申请人比另一个申请人更有可能违约，而且他们违约的可能性是另一个申请人的2倍。

But another thing it has to do is it has to figure out the absolute levels of default risk. So that first baseline borrower, what does a nearly risk free borrower look like? What does a super risky borrower look like? And what is the average population default rate going to look like?
但它还需要做的另一件事是，它必须确定违约风险的绝对水平。因此，第一个基线借款人，一个几乎没有风险的借款人是什么样子？一个超级风险的借款人是什么样子？平均人口的违约率将会是什么样子？

And the reason that those are such separate problems is that for that second problem, you really have, in some sense, very limited amounts of historical data to train on in the way that you would like to train an ML model. Whereas for the first problem, the risk separation problem, you have an enormous amount of data that you could look at basically every single default that's ever happened is mostly a story of relative default.
那些是分开的问题的原因是，针对第二个问题，你实际上只有非常有限的历史数据可以用来训练机器学习模型。而对于第一个问题，风险分离问题，你有大量的数据可以查看，基本上每个违约事件几乎都是相对违约的故事。

It tells you that within a given period of time, how likely was a particular borrower or a historic default compared to some other borrower characteristic? Whereas to look at the average level of default in say a year, well, you only have so many years on record and you certainly don't have the thousands or millions of data points that you would like to have.
它告诉你，在给定的时间段内，一个特定的借款人或者历史违约比其他借款人特征更有可能发生违约。而如果只是看一年内的平均违约水平，那么你只有有限的记录年份，在数据点方面肯定没有你想要的成千上万个数据点。

And human history barely goes back that far. And certainly we weren't recording our loan repayments going back to those days. So that makes it very challenging. And so we sort of have just like a handful of meaningfully different periods, recessions and non recessions, you can think of it as like maybe like 20 data points in recorded history. And 20 is just not enough for any kind of serious machine learning training problem.
人类历史其实并没有那么久远，而且我们在那些日子里也没有记录我们的贷款还款情况。所以这使得问题变得非常具有挑战性。我们只有少数几个具有实质不同的时期，比如经济衰退和非经济衰退时期，你可以把它想象成历史上大概只有20个数据点。而20个数据点对于任何严肃的机器学习训练问题来说都不足够。

And so we separated this sort of micro question and this macro question. And we said, okay, how can we just precisely measure the macro component and know whether a period of time is seeing a increase or decrease and of how much compared to past periods in its average level of default across the population?
所以我们把这个微观问题和宏观问题分开来。我们说，好的，我们怎样才能精确地测量宏观部分，并知道一个一段时间是否在整个人口的违约率平均水平上出现了增加或减少，以及相对于过去时期的增加或减少多少呢？

So that's what we wanted to know. And about 18 months ago, we started pouring a lot of our research science attention to this problem. And the end result is something that today we call the UMI, the upstart macro index.
那就是我们想要了解的。大约18个月前，我们开始将大量研究科学的关注投入到这个问题上。最终结果就是今天我们所称之为UMI的东西，即新兴宏观指数。

Yeah, the upstart macro index. Essentially, it takes a, it aims to tell you if you had constant borrowers. If you were just constantly making loans to the exact same group of people across different periods of time, what would the delta in their delinquency rates be in different days, weeks, months, years, whatever timeframe you want to compare?
是的，这是一个新兴的宏观指数。它的基本原理是告诉你是否有常常借款的人。如果你在不同时间段内一直向同一组人借款，那么在不同的日期、周、月、年或者任何你想比较的时间范围内，他们的拖欠率会发生怎样的变化？

And the way it does this is by making use of what we call a reference model. Because of course, you can't actually find exactly identical borrowers every single year to make loans to. And no matter how you control it with one variable, two variable, you might say, well, let's just look at borrowers who are 680 to 720 on the FICO scale.
它这样做的方式就是利用我们所称的参考模型。因为当然，你不能每年都找到完全相同的借款人来贷款。无论你如何使用一个变量或两个变量来控制，你可能会说，好吧，让我们只看FICO评分在680到720之间的借款人。

That would be a very simplified way to do this. But of course, there are a million other things to know about these borrowers other than their FICO score. And so the way to do that in a more holistic and continuous way is with what we call a reference model.
那就是一种非常简化的做法。当然，除了信用评分外，还有无数其他和借款人相关的事情需要知道。所以更全面和连续的方法是使用我们所谓的参考模型。

So you say, what if we under wrote all of these loans with exactly the same model? Because of course, in reality, every loan you have on your books is underwritten with probably a different model at a different period of time. But if you instead under wrote them with exactly the same model and made predictions for every single month that you've subsequently observed, how do the actual delinquency rates and default rates compare to what a constant reference model would have predicted?
"你说，如果我们用完全相同的模型承担所有这些贷款会怎么样呢？因为实际上，你账面上的每笔贷款可能都是在不同的时间采用不同的模型进行承保的。但如果你改为使用完全相同的模型进行承保，并为后来观察到的每个月份做出预测，那么实际拖欠率和违约率与常态参考模型相比如何呢？"

And what that tells you, the reference model is doing the work of controlling for the consumer and they what you can observe is, of course, the actual loss rates in real calendar months compared to what the model would have predicted on this sort of abstracted hypothetically constant kind of calendar time period.
这句话的意思是，参考模型正在控制消费者，而您能够观察到的当然是在现实日历月份中的实际损失率，与该模型在抽象的虚拟日历时间段中预测的相比。

And so the deltas in those observations compared to a sort of now controlled expectation of default or delinquency is going to tell you how much more or less risky 2022 was compared to 2021 compared to 2019 compared to 2017. And so that's what we've developed. We call it the absurd macro index and it does a really good job for us of both explaining how a sort of past delinquency patterns have played out and why they've played out that way.
因此，这些观察值中的增量与现在控制的违约或拖欠预期相比会告诉你，2022相对于2021、2019和2017更或更少有多大的风险。这就是我们所开发的东西。我们称之为荒谬宏观指数，它对于解释过去拖欠模式的发展和原因有很好的作用。

And also actually in being able to make forward looking statements of the forum, well, if the macro environment gets 50% worse, then this is what will happen to your loss rates and being able to make statements of that form, of course, then allows us to make useful statements about how we should be building in buffer and conservatism to the underwriting for any given lending partner, how aggressive or conservative their parameters are set in how much lending they want to do.
实际上能够对论坛做出前瞻性声明也是很重要的，如果宏观环境恶化了50％，那么你们的损失率会发生什么变化。能够做出这种形式的声明使我们能够在为任何特定的放贷伙伴建立商业政策时，做出有用的声明，包括我们应该如何建立缓冲区和保守主义，以及他们的参数设置有多么激进或保守，他们想要进行多少放贷。

So a really useful both backwards looking and forwards looking. So I got a number of questions to dive off of this one of the things you said. But first I kind of want to be clear, when you produce the index, it's a singular number for any given month as I understand it. How should someone look at the number and say, what does this tell me if it's a 1.0, a 2.0, a 0.5? Like what's the actual index and how do I think about reading it? And then I want to dive in a bunch of things that you said about this.
所以它其实非常有用，无论是回头看还是前瞻性。所以我有很多问题要从您之前提到的内容中展开。但首先我想要弄清楚，当您编制指数时，我理解它是每个月的一个单一数字。如果是1.0、2.0、0.5之类的数字，某人应该如何解读它，这个指数实际上包含了什么信息，我该如何思考它？然后我想深入探讨一些您提到的内容。

Yeah. So the index is condensed down to a single number, as you just said. And that single number can be applied to any length of time. So usually we talk about it in one month periods because most of our loans have one month repayment period. So it makes a lot of sense to look at it that way. But you could look at them in days, weeks, years, quarters, however you like. But for a period of time, you're going to have a number that number is going to range all of the between 0 and 3 generally. Though, of course, there's no theoretical upper limit to how high the UMI could go.
嗯，所以指数被压缩成一个单一的数字，就像你刚才说的那样。这个单一的数字可以适用于任何长度的时间。通常我们讨论的是一个月的时间段，因为我们的大部分贷款都有一个月的还款期限。所以以这种方式看待它是很有意义的。但是您也可以根据天数、周数、年份、季度等来看待它们。但在一段时间内，您将会有一个数字，这个数字通常在0和3之间波动。当然，UMI可以达到的最高理论数值没有上限。

But what we've normalized the numbers to mean is that as a 1.0, generally is meant to correspond to a, what we think of as a sort of historically normal time, or at least historically normal time in our reported history. What you can generally think of as being the sort of five years before COVID. And those five years, there was a relatively sort of a stable macro environment. We do not see large changes, you know, taking that reference model approach. If you're just underwritten all the loans using the same model, you would basically see the same sort of results across the different calendar years.
我们对数字的常规化意义是，1.0通常意味着对应于我们所认为的历史上正常的时间，或者至少是我们所报道的历史中正常的时间。你可以想象一下，在COVID之前的五年中，这被认为是比较稳定的宏观环境。在这五年里，我们并没有看到太大的变化，如果你采用参考模型的方法去批准所有的贷款，你基本上会在不同的日历年份得到相同的结果。

Whereas, whereas after that, of course, once you get into the COVID world, you see both large positive and negative shocks to that, which is exactly what this methodology is hoping to tell us about. Yeah, imagine that the COVID period, particularly the very serious economic impact than the rapid input of stimulus, and then the rapid withdrawal of stimulus was like changing of macro conditions on steroids, or at least at fast forward speed in terms of a normal cycle, which I think is kind of what was some of the impetus for this is that kind of rapidity of change. What we usually think of is maybe not a super slow, but not as rapid of a changing environment as we experience over the last 24 months.
当你进入COVID的世界之后，你会看到对经济产生了积极和消极的冲击，而这正是这种方法希望告诉我们的。想象一下，COVID时期，特别是经济影响非常严重，超过了快速注入刺激和快速撤回刺激，就像类固醇影响宏观环境的变化，或者至少是以正常周期的快进速度进行变化，我认为这是一些促成这种情况的原因，因为我们通常想象的环境变化可能不会那么缓慢，但也不会像我们过去24个月经历的那么迅速。

Yeah, totally. I mean, it was almost a perfect couple of years to showcase just the difficulty of the macro prediction problem, which we've always felt was not really a good prediction problem in the sense that these, you have so few of these events as we talked about, and each shock looks really different. So trying to predict them based on past shocks is what's kind of silly. And I mean, April 2020 was probably the most extreme example of this, where April 2020 hits. And of course, we have built over the years plenty of what you'd call macro prediction models that look at things like unemployment rates and various sort of macro indicators that unemployment generally has in past shocks been the single biggest driver.
是啊，完全没错。我是说，这几年几乎是完美的例子来展示宏观预测问题的难度，我们一直感觉这并不是一个真正的好预测问题，因为像我们谈论的那样，这类事件很少，每次冲击看起来都非常不同。因此，基于过去的冲击来预测它们就有点儿愚蠢。而且，我的意思是，2020年4月可能是最极端的一个例子，4月份来袭。当然，多年来我们已经建立了许多你会称之为宏观预测模型的东西，看一些东西，比如失业率和各种宏观指标，过去的冲击中，失业率通常是单一最重要的驱动因素。

But what happened in April 2020, while unemployment rates went through the roof, I sort of never before seeing rapid rise in unemployment. And yet what happened to low-repayment, well, it turns out that within a very short time low-repayment did not only do not get worse, it became the best period of low-repayment probably ever in all of human history. And it was because, of course, that the government responded very forcefully to the increased unemployment pandemic and did a lot of stimulus in various ways. And people had a lot of extra cash to make their loan payments and not as many places spend it because they were locked up at home.
2020年4月发生了什么事情呢？失业率飙升，让我见证了前所未有的失业激增。但是，对于低借款率，发生了什么呢？事实证明，在非常短的时间内，低借款率不仅没有变得更糟，而且可能是人类历史上最好的低借款期。这是因为，政府对失业疫情做出了非常有力的反应，并在各种方式上进行了很多刺激。人们有了很多额外的钱来偿还贷款，因为他们被困在家里，没有太多地方可以去消费。

And so you got exactly the opposite effect of what sort of historical precedent what I've told you. And then if you look to 2022, you get almost the mirror image of that where you have extremely low unemployment, basically pushing records in terms of how low unemployment gets. And yet you have this rapid rise in delinquency rates and it's for exactly the same reason. The stimulus ends. People still had jobs, but it turned out that printing money and giving it to people had a bigger impact on whether they would repay their loans and seeing if they were working hard on it.
所以你实际上得到了完全相反的效果，就是我告诉你的历史先例。如果你看到2022年，你几乎可以看到镜子中的画面，极度低的失业率，基本上刷新了失业率的最低记录。然而，由于同样的原因，你看到拖欠率迅速上升。刺激结束了。人们仍然有工作，但经验证明，印钞和将钱给人们对他们是否偿还贷款以及他们是否认真付出更大的影响。

So I want to ask about how traditional lender, I mean, this is certainly the topic of how is the macroeconomy playing into loan performance? Every time I talk to lenders are looking at inflation rates and interest rates and unemployment rates and trying to understand what that's going to mean or is doing to a given portfolio. So how do you think traditional lenders have looked at the kind of problem you're trying to model or at least the data point of like what is the macro environment doing to the loss rates in a pool? And what are the weaknesses in the way it's traditionally been done that you were trying to solve for when you thought about building out the UMI methodology?
我想问一下传统贷方，也就是说，宏观经济对贷款表现的影响是一个非常重要的话题。每次我和贷方交谈时，他们都会关注通货膨胀率、利率和失业率，试图了解这些数据对一个给定组合的贷款损失率会产生什么影响或者已经产生了什么影响。那么，您认为传统贷方是如何看待您正在尝试建模的问题，或者说至少从宏观环境角度考虑，贷方如何看待贷款池中损失率的数据点？在传统方法中存在哪些弱点，您在思考如何构建UMI方法时试图解决的问题是什么？

Yeah, so I think there's spectrum of approaches, but they share a couple things in common.
嗯，我认为有不同的方法，但它们有几个共同点。

The first is that because many traditional lenders don't change their underlying models very much or they have very simple ones that have a very limited number of variables, they can kind of make an assumption.
首先，由于许多传统的贷款机构在其基础模型上并没有进行太多更改，或者它们所使用的模型非常简单，只有非常有限的变量，因此它们可以作出一些假设。

And it doesn't seem as crazy as it does in our world though, we'll get to why it still suffers from some of the same challenges, but it doesn't seem as crazy to make the assumption that hey, we are underwriting with a constant model and so the borrowers that we produce across different years should roughly be the same and if they're roughly the same, then I can understand where the macro is just by looking at the average default rate in my own portfolio.
在我们这个世界里，这种想法似乎并不像看起来那么疯狂，虽然它仍然面临一些相同的挑战，但这个假设并不像疯狂那么荒唐：我们正在使用一种恒定的模型进行承保，因此我们在不同年份生产的借款人应该大致相同。如果他们大致相同，那么我可以通过查看自己组合中的平均违约率来理解宏观经济状况。

And this is sort of okay. It's sort of okay because if you actually don't change your model in any way, then actually you do have a fairly constant model.
这有点可以接受。因为如果实际上您不以任何方式改变您的模型，那么您确实拥有一个相当不变的模型。

But of course it also assumes that your borrower mixes on changing across time that your pricing approaches have not changed across time and actually hasn't changed in any way across time and of course we know that even if you just were using the FICO score done, and of course the FICO score distribution is changing across time and the inputs to it are getting changed by how medical bankruptcies are handled and different things like this.
当然，这也假设你的借款人在时间推移中交错变化，你的定价方法在时间推移中没有改变，实际上在时间推移中没有任何改变，当然，我们知道即使只使用FICO评分，FICO评分分布也会随着时间的推移而变化，它的输入也会受到如何处理医疗破产等不同事项的影响而发生变化。

So there are actually a bunch of these changes happening under the hood and maybe they're not that extreme, but put together they're probably each 10-20% type effect.
其实在背后会有许多这些变化，也许它们并不是那么极端，但加在一起，它们可能会产生10-20%的影响。

And soon you could be in this world where you might be seeing default rates that are 50% higher and now you have to ascribe that to macro.
很快你可能会进入这样的世界，其中违约率可能会高出50％，现在你必须将其归因于宏观经济因素。

When in reality it might just be that your borrower makes change 20% and your underwriting model will change 20% and there you are with the 50%.
实际上可能只是你的借款人做出了20%的改变，而你的核保模型也会相应改变20%，这就让你达到了50%。

So that's pretty inherently challenging to measure macro in.
所以，这对于衡量宏观因素来说是非常困难的。

And then I think on the sort of response side to macro, I think it's also when you don't have a precise factor that's able to separate what your underwriting model is doing and what is happening in the macro.
然后我认为在宏观层面的回应方面，当你没有一个能够分离你的承保模型正在做什么和宏观情况的精确因素时，也会出现这种情况。

There's always this temptation to say, well, we can fix the problem just by changing the underwriting model.
我们总是忍不住想说，我们只需要改变承保模式就可以解决问题。

So if we only focus on higher FICO borrowers or higher income borrowers then we don't have to worry about the effects of a recession or the effects of the macro shock.
如果我们只关注FICO信用评分高的借款人或收入高的借款人，那么我们就不必担心经济衰退或宏观冲击的影响了。

And I think that is an approach that really can get you into trouble where you look at a lot of past recession something like 2008 and you look at what happened to credit card delinquency rates in 2008 or more insurances in 2008 and what you see is actually that the largest multiple increases to delinquency and default rates were in high FICO buckets.
我认为这种方法可能会让你陷入麻烦，当你看到很多过去的经济衰退，比如2008年，你看看2008年信用卡拖欠率或更多的保险，你会发现实际上拖欠和违约率最大的倍数增长出现在高FICO分数档案中。

And maybe it's a little counterintuitive but then if you think about what was happening in 2008 of course one thing is that the biggest stress were on people who were buying large homes and had big, big mortgages and of course that's going to be a primer consumer.
也许这有点违反直觉，但是如果你想一想2008年发生了什么，当然一个情况是那些购买大型住房和负有巨额抵押贷款的人承受的压力最大，当然，这些人是主要的消费者。

And another thing is that of course it matters quite a lot, the kind of shock in the economy and whether that's something that's going to impact more prime people or less prime people.
还有一件事就是，当然了，这是相当重要的，经济的冲击类型和它是否会对更多的优质人才还是次优质人才产生影响，这是非常关键的。

But some of the less prime people actually in some sense are like always in a recession. They're kind of living paycheck to paycheck in a way that means their lost rates are normally much higher but it's also harder for them to get so much higher.
有些不那么优秀的人在某种程度上就像经常处于衰退状态。他们过着靠着薪水支撑生活的生活方式，这意味着他们的失业率通常要高得多，但对于他们来说，要提高他们的收入水平更加困难。

Whereas when you have someone who goes from being comfortably putting money in the bank every month to suddenly lose their job, it's a fairly tremendous change in their ability to pay things and of course a debt obligation is much larger.
当一个人从每个月轻松地往银行存钱转变为突然失业时，他们还款的能力会发生非常巨大的变化，并且债务义务也变得更加庞大。

There's less ability to easily make up for it just by going to backup sources of cash. There's these different effects that play out.
仅仅通过去备用的现金来源来轻松弥补它的能力变少了。有这些不同的影响会产生作用。

It's very hard, I think, nearly impossible to solve for macro problems by changing the borrower level characteristics that you underwrite to because there are no set of characteristics that are inherently recession proof or macro shock proof.
我觉得，要通过改变你评估的借款人水平的特征来解决宏观问题非常困难，几乎是不可能的，因为没有一组特征是天然的抗衰退或宏观冲击的。

And even if there were frankly, I think we don't realistically have the historical data to identify what those are today.
就算有，说实话，我觉得我们现在没有历史数据来识别它们。

The other thing that occurred to me as I was reading through the approach to UMI is the sense that particularly to see things in real time, what you've really said is to do this right, I need to wait a whole the borrower-based constant, some sort of a constant reference model, the credit model constant.
当我阅读处理UMI的方法时，我意识到另一件事，那就是特别需要实时观察事物，你真正说的是要做到这一点，我需要等待一个借款人为基础的恒定参考模型，即信用模型常数。如果有必要的话，请改写。

But I also need really an individualized risk prediction and ideally a month-based risk prediction because if I want to see what happened in this month, I want to see that I expect this some a month or every given month of seasoning that could be for your loan portfolio exists in the portfolio somewhere.
我也需要一个真正个性化的风险预测，最好是基于月份的风险预测，因为如果我想看看这个月发生了什么，我希望看到我预计这个月或每个赛季中存在于你的贷款组合中的某些贷款组合存在于组合中的某个地方。

You want to know, is this loan that went to link when a month 32 was that unexpected or expected?
你想知道的是，这笔贷款在一个月32将到期，这是出乎意料的还是预期的？

So you need really a pretty precise model of risk to be able to see this in something like real time, I mean not real time, but to see it rapidly and quickly and to be able to adjust. And so it occurred to me that the kind of individual risk prediction and loan level and month level risk predictions in the reference model are really critical to being able to see not just with granularity but with speed, what's actually happening in the environment and how it might be changing than the underlying from what the underlying predictions might have expected.
所以，你需要一个相当准确的风险模型，才能像实时一样快速地看到这一点，我不是说真正的实时，而是能够快速地看到并调整。因此，我认为参考模型中个体风险预测和贷款级别以及月度风险预测的种类对于不仅仅是以细粒度，而且是以速度来看到环境中实际发生的事情以及它如何可能会从基础预测所期望的情况中发生改变非常关键。

Yeah, that's totally right. If you only have lifetime default probability predictions or you only have those for certain buckets of borrowers and you're not going to be able to do this thing, you won't be able to control for the borrower mix at all. With a reference model that doesn't consider enough individual characteristics. And if you only have a lifetime prediction, then of course you can't compare what's going on in different periods of time. So you really need a model where you have a lot of confidence that the month seven prediction for this particular individual really is a good prediction, at least in relative terms relative to your predictions in the other months and relative to your predictions for other borrowers in that month. If you have confidence in that, maybe think of as if there are months on the x-axis and borrowers on the y-axis, you need to feel like both your axes are really doing a lot of work for you that it's sort of an on-bias estimate. And if that's the case, then you can actually take that point on the matrix if you want to compare it to other points on the matrix and it tells you something meaningful. And of course, if you don't have that, then you can't do that comparison.
没错，完全正确。如果你只有借款人终身违约概率的预测，或者只有针对某些借款人群的预测，你就无法控制借款人群的混合。而且，如果没有考虑足够多的个人特征的参考模型是无法做到的。如果你只有终身预测，那么当然你就无法比较不同时间段发生的情况。因此，你真的需要一个模型，以便你确信每个月的第七个预测对于这个特定的个体来说确实是一个好的预测，至少在相对于其他月份的预测和相对于该月份其他借款人的预测方面是好的。如果你对此有信心，也许可以将月份放在X轴上，借款人放在Y轴上，你需要感觉到两个轴都在为你做很多工作，它是一种无偏估计。如果是这种情况，那么你实际上可以将该矩阵上的点与矩阵上的其他点进行比较，这会告诉你一些有意义的信息。当然，如果你没有这个，那么你就无法进行比较。

Yeah. So I did want to ask, you know, you call this the macro index. You talked about the macro and micro prediction levels of the model. And it's interesting because macro can be a reference to the macro economy, the kind of broader world, but it's also kind of a size metric macro and micro. And so I kind of want to ask the question, like how do you think about the relative importance in any given moment or overall for the micro portion of what you might call the core risk modeling that that upsides being focused on? And it's kind of macro level that's maybe harder to predict. So now, somewhat more measurable through UMI. How do those kind of stack up in terms of how important they are to a lender into a lender's portfolio performance?
嗯，我想问一下，你会把这叫做宏观指标。你谈到了模型的宏观和微观预测水平。这很有趣，因为“宏观”可以指宏观经济，更广泛的世界，也可以是大小指标的宏观和微观。所以我想问一个问题，你如何考虑任何时刻或总体上所谓的核心风险建模中微观部分的相对重要性？而在宏观水平上，也许更难预测。现在，通过UMI可以测量得到。对于贷款人和贷款组合的表现来说，它们在重要性上如何排名？

Yeah. I might first say they're both really important. They're both important in the context of, you know, fixed income of any kind has this property that you can lose a lot more money than you can make. That's the nature of fixed income. And that means, you know, generally anything that can meaningfully increase your default rates is something you're really going to care about. And certainly both the macro and the micro here fall into that bucket. If you get those really wrong, it's going to really affect the return you were expecting.
嗯，我可能会先说这两个都非常重要。在任何固定收益的背景下，它们都显得非常重要，因为这种类型的投资存在比预期亏大得多的风险。这就是固定收入的本质。这也意味着，你会真正关注能够显著提高违约率的任何东西。而这里的宏观和微观方面都落在了这个类别中。如果你对这两个方面都完全错了，那么它将会极大地影响你预期的回报。

Now, having said that, if you were to just force a competition between the macro and the micro, how much can loss is very based on macro and how much can loss is based on micro? I mean, frankly, then it's not even close because the macro, you know, you can get, you can get a period of time that's three or four, maybe in an extreme case, five times as bad as another period. In micro, of course, you can get borrowers, not just theoretically, but even identifiably, for example, in our model, you can identify borrowers who are like 80% likely to default, and you can identify borrowers who are 0.8% likely to default. So you're talking about a 100x spread between the highest and lowest risk in micro, and you're talking about a 5 to 1 spread in the highest and lowest risk in macro. And so, of course, it's so much more important to get the micro right than the macro right.
那么，如果你只是强行在宏观和微观之间进行比较，损失有多大是基于宏观，有多大是基于微观？说实话，宏观和微观之间的差距是非常大的，因为在宏观上，你可以遭受到比另一个周期多三到四倍，甚至在极端情况下高达五倍的危险期。在微观上，当然你可以找到那些不只是理论上，而且还能够明确鉴别的借款人。例如，在我们的模型中，你可以找到那些可能有80%违约风险的借款人，同时也可以找到那些仅有0.8%违约风险的借款人。因此，在微观上，最高和最低风险之间的差距可以高达100倍，而在宏观上，这个范围只有1比5。所以，当然，把微观掌控好比正确地掌控宏观更加重要。

There are so many more ways that even in the best of times, if you try to go to market and underwrite with an extremely poor performing model, just everything will go wrong. Because not only because your model will be wrong, you face adverse selection, because other vendors will pick off at lower APRs, the better borrowers, and just so many things can go wrong in micro, whereas in macro, things can also go wrong, such as, you know, when you go from a really, really unreasonably good period, like 2021, to really sort of sharply bad period, like 2022.
有很多更多的方式，即使在最好的时候，如果你试图去市场并用一个极差的执行模式来承保，一切都会出错。因为不仅是因为你的模型会出错，你还会面临逆向选择，因为其他供应商会以更低的APR挑选更好的借款人，在微观层面上有很多事情可能会出错，而在宏观层面上，事情也可能会出错，比如你从一个非常不合理的好时期转变为一个非常急剧的不好时期，比如2021年到2022年。

And that's going to significantly and negatively impact lender returns, but it's still going to be on, you know, if we're talking about annualized return impact, I mean, depending on what your risk mix is, and you're still talking, you know, you can count it on your two hands, like the number of percentage points of impact that's going to have, whereas if you get the micro wrong, I mean, there's really no limit to how much money you can lose doing this, and of course, on the flip side, you know, when things are going well, micro is what enables you and has enabled the lenders that use our platform over the years to be able to make dramatically more income running a lending business, because micro is what enables you to confidently expand your approval box while holding your risk levels constant.
这会对贷款人造成显著而负面的影响，但如果我们谈到年化回报的影响，这种影响还将保持在一定范围内。这取决于你的风险组合，你可以把它算成几个百分点，不用全部手掌都数。而如果你的微观风险估计出现了问题，你的损失就无限制了。当然，在情况良好时，微观风险评估是促使我们的平台上使用的贷款人能够通过运营贷款业务获得大量收入的关键，因为通过微观风险评估，你可以在保持风险水平不变的情况下，自信地扩大你的批准范围。

So if you can improve 50% or 100%, or 200%, as many borrowers as you previously could, but you get the same risk levels, then of course, you know, something now you're able to have meaningfully change the amount of lending and helping generated.
如果你能把贷款对象的数量提高50%、100%或200%，但风险水平不变，那么显然，现在你能够有意义地改变贷款和帮助的数量了。

Yeah, interesting. Well, I want to dive in a different topic. You made this comment about like if you can model out if the world, the macro got 50% worse, or I'd love to say 50% better.
嗯，有趣。好的，我想要转入另一个话题。你曾发表过这样一条评论，如果可以模拟出世界，宏观经济会变差50%，或者说我很希望说变好50%。

Sometimes I worry that we talk about the macro. We only talk about bad macro, I'm talking about macro getting worse and protecting, not the fact that particularly from where we sit, as we're recording this, there's a lot, probably more upside for the macro getting better than substantially worse over the next at least 12 or 18 months.
有时我担心我们只谈论宏观经济。我们只会谈论不好的宏观经济，我说的是宏观经济在恶化并得到保护，而不是从我们录制这段视频的角度来看，宏观经济未来至少12至18个月内会好过恶化。可能会有更多积极的方面。

But the interesting thing is you haven't talked really about what I would consider traditional macro variables, right? Like most people when they talk about how is the macro doing, you know, you have a UMI, you'll talk about where it's at. But most of you look at inflation, unemployment, interest rates, maybe savings rates, things of this nature that are very clearly tied to what we think of as a macro economy GDP or something like that.
但有趣的是，你们还没有真正谈论我认为传统的宏观经济变量，是吧？就像大多数人谈论宏观经济表现时，你们可能提到UMI，谈论它的状态。但你们大多数人观察的是通货膨胀、失业率、利率、储蓄率等类似的经济指标，这些指标与我们所认为的宏观经济GDP等相关。

Is there a relationship or those part of the UMI model, talk to me a little bit about how like traditional views of what is happening in the macro relate to the upstart macro index or how you think about the impact of macro and the lending portfolio?
请问UMI模型中是否存在关系？您能否稍微谈谈传统观念如何与新兴宏观指数相关，或者您如何思考宏观和贷款组合的影响？

Yeah, they're super correlated in short. The upstart measured UMI, we like it because it has the nice property that it's directly applicable to the math you want to do on lending. And because it's really specifically designed to tell you this period is a period in which you can expect twice as many dependencies on the same bar or pool as that period, you know, it lets you do math in a very precise and direct way.
是的，它们之间有很强的相关性。我们喜欢UMI这个新生代的度量方式，因为它有一个很好的特点，就是能够直接应用于你希望在贷款上做的数学计算。而且，因为它被专门设计成能够告诉你，在这个时间段内，你可以预计同一个酒吧或池塘会有两倍的依赖关系，所以它让你能够以非常精确和直接的方式进行数学运算。

But of course, it would be really concerning if this index that's meant to measure the macro is totally unrelated to the sort of traditional macro economic variables, which of course, kind of though each one may be able to tell a part of the story, they are collectively, you know, a pretty good description of the things that we care about in the economy.
当然，如果这个旨在衡量宏观经济的指标与传统宏观经济变量完全无关的话，那真的是非常值得关注的。虽然每个变量可能只能讲述故事的一部分，但它们作为一个整体，就能够相当好地描述我们关心的经济状况。

And so we have that's been an important part of developing the UMI, is actually building a sister model that tells us what we haven't found a nice sort of externally marketable name for it yet. But you think about there's sort of UMI as measured in the sort of collective repayment rates of the of all upstart platform originated loans where we have access to the data. But then there's UMI as you would predict based on what you're seeing from the sort of official macro economic data, the unemployment rate, the inflation rate, the personal savings rate.
所以，我们开发UMI的一个重要部分是建立一个姐妹模型，告诉我们还没有一个好的对外销售的名字。你可以想象一下，UMI是通过我们获得的所有起点平台贷款的集体偿还率来衡量的。但是，UMI也可以根据官方宏观经济数据，如失业率、通货膨胀率和个人储蓄率来预测。

And what you want is you want to see that these two measures, they sort of internally measured UMI and the externally inferred UMI are largely moving together, highly correlated and kind of moving together. And that's exactly what we see, we've been able to see that.
你想要的是希望看到这两个度量，内部测量的UMI和外部推测出的UMI基本上是一起移动的，高度相关并且一起移动。我们确切地看到了这一点。

And actually, you can, you know, the maybe short and snappy version of this is if you just want what single variable released by the Fed can tell you roughly where the UMI is going. And the answer would be the personal savings rate.
实际上，你知道的，这个简短而又简明的版本是，如果你只想知道联邦储备委员会发布的一个单一变量大概可以告诉你美国经济的大致走向，那就是个人储蓄率。

And that of course is a bit cheating because the personal savings rate itself is a bit of an amalgamation of a bunch of variables. It's, you know, looking at personal, probably looking at personal income and personal expenditures. But of course, what makes up personal income, you know, people have to work. So there's employment rates baked into that. There's wage and wage growth baked into that.
当然，这有点作弊，因为个人储蓄率本身就是一堆变量的混合体。它在考虑个人收入和个人支出，但是当然，组成个人收入的因素中，你知道的，人们必须工作。所以其中会涉及到就业率，工资和工资增长。

And of course on the personal expenditure side, there's sort of your real rate of spending and then there's the inflation rate. And so you kind of got this one variable that Fed publishes, you know, the end of every month and it tells you roughly the things that really matter, unemployment, inflation, spending and wage growth. And the net of that sort of is like how much money people have left over, which is a really good proxy for, you know, how much money people are going to have available to service their debt.
当然，在个人开支方面，有实际开支率和通货膨胀率两种。因此，美联储每月发布的一项变量涵盖了真正重要的事项，如失业率、通货膨胀率、开支和工资增长率。这个变量的净值就是人们余下多少钱，这是衡量人们还有多少钱可以用来偿还债务的一个非常好的标准。

And if you just look at that one variable, that's probably like, you know, mostly, mostly correlated with how we've measured UMI over the past years. It's a, yeah, it does seem like a I won't say a cheating variable, but it is an amalgamation that takes into account a number of the things that people traditionally think of as pretty impactful to at least, particularly unsecured credit where there's, you know, it's kind of the, I won't say the riskiest, but the bottom of the payments act in many ways where the first thing to take a, take a hit one time to get tough.
如果你只看那一个变量的话，那大概是，你知道的，大部分的，大部分与我们过去几年里衡量UMI的方式相关。它看起来是一种综合变量，考虑了人们传统认为对至少，尤其是未担保信贷产生相当重要影响的几个因素。它是一种，是的，我不会说是舞弊变量，但从许多方面来说，它考虑了许多东西，尤其是在未担保信贷方面，那里有，你知道的，它可能不是最高风险的，但在付款行为中它属于底部，在一次遭受冲击时它是最容易受到打击的。

So one thing we haven't talked about today at all, and I'm sure our lending partnership seems to be very upset with me to be this long to ask this question. But, you know, I think there's obviously, when UMI goes up, if that macro gets riskier, there's this question of, is that a direct indication of loans under or overperforming?
今天我们还没有谈到的一件事，我敢肯定我们的贷款合作伙伴似乎对我提出这个问题感到非常不满。但是，你知道，当UMI上涨时，如果这个宏观形势变得更加风险，那么就会有一个问题，那就是是否直接表明贷款表现好坏？

And how does the sense of the macro play into how you actually work with a lending partner to assign a level of risk to a borrower at time of underwriting and understand the returns produced by that portfolio? So I think the thing that I hear a lot is like, does a UMI above one, meaning all my loans are underperforming because the macro is high and therefore I'm losing money and that's a bad thing and I should stop lending or whatever.
"那么，宏观感觉如何影响你与贷款伙伴合作，在核保时为借款人分配风险水平并理解该投资组合产生的回报？我认为我经常听到的问题是，如果UMI（不良贷款占比）超过1，这是否意味着我所有的贷款表现不佳，因为宏观环境不佳，因此我会亏损和这是一件不好的事情，我应该停止出借或做出其他选择。"

So talk to you about how you leverage this in the context of partnerships with the originators to kind of calibrate to what's going on and help them understand the level of risk in the environment and that this going into their originations.
所以，跟你谈谈你如何利用这一点，在与原始资产方合作的背景下进行校准，帮助他们了解环境中的风险水平，以及这会进入他们的资产原始管理流程中。

Yep. Yeah, so the insure, no, it does not, there's no direct way to say UMI is a certain number where therefore my loans are either over or underperforming to know that you need to know one more thing, which is essentially what UMI were you assuming when you made the loans in question. And that particular thing we call the UMF, the macro forecast.
是的。嗯，关于保险，不，没有直接的方法去说UMI是一个确定的数字，因此我的贷款要么表现好要么表现不佳，要知道这一点，你需要知道另一件事，那就是你在做出贷款决策时所假设的UMI。而那个具体的东西我们称之为UMF，也就是宏观预测。

And if you very roughly speaking, if you just for any batch of loans, if you take their UMF and you then look forward and say, okay, I made this batch of loans in January of 2022 and at the time we had a UMF of 1.0 and now the UMI is 1.5, then that sort of implies that you're going to have actual losses that are 50% higher than you forecast.
如果你简单地说，如果你只是针对任何一批贷款，拿他们的UMF然后向前看，比如说，你在2022年1月份放出一批贷款并且在那时UMF是1.0，现在UMI是1.5，那么这暗示着你将会实际损失比你预测的要高50%。

And in that case, sort of example, you gave as valid. Now fast forward to more recent times and many of our lenders actually have a much higher UMF, imagine your UMF is 2.0, but the UMI that's been observed since is only 1.8. Well then actually, even though 1.8 is a really high number by historical standards, it's actually 20 or it's actually 10% lower than the UMF that you assumed.
在这个例子中，你给出了一个有效的例子。现在快进到更近期，我们的许多贷方实际上拥有更高的UMF，想象一下你的UMF是2.0，但自那时以来观察到的UMI只有1.8。那么实际上，即使从历史标准来看，1.8是一个非常高的数字，但它实际上比你假设的UMF低了20%或10%。

And so that means that your loans were priced as if you were operating in a 2.0 type world, but actually you've been seeing a 1.8 type world and now you're going to see defaults and delinquencies that are running on average 10% lower than you were expecting, which of course means more net interest, more access returns and so returns will be higher than expected.
这意味着你的贷款是以2.0类型的环境定价的，但实际上你正在经历一个1.8类型的环境，现在你将看到的违约和拖欠平均要比你预期的低10%，这当然意味着更多的净利息、更多的净收益，因此收益将高于预期。

And of course, it actually matters how this plays out over the whole term of the loan. But just roughly as an approximation, you think about a single point in time, this kind of tells you this ratio between UMI, UMF tells you how things are going to be. And so if you underwrite with a fairly high UMF, then of course there's more room for UMI to positively surprise against that.
当然，在整个贷款期限内这种情况如何发展确实很重要。但是仅就一个时间点而言，这种UMI和UMF之间的比率可以大致告诉你事情的走向。因此，如果你使用相当高的UMF进行核保，那么肯定有更多的空间让UMI对其产生积极的惊喜。

On the other hand, if you underwrite with a high UMF, you're going to be passing up a lot of borrowers and pricing a lot of other ones out of converting just because you are assuming that the world is in a really bad place and you want to lend to people who are going to be going through a really hard time. And of course, if they aren't or if other lenders are not seeing it that way, then those borrowers are going to go elsewhere.
另一方面，如果您选择高的UMF来承保，您将错失许多借款人，也将因为您假设世界面临着巨大的困境，只想贷款给那些将会经历非常艰难时期的人，而导致其他很多借款人失去了融资的机会。当然，如果那些借款人并没有处在那样的情况下，或者其他贷方并不认为情况如此，那么这些借款人就会去寻找其他渠道。

How can you help a lender think about, I mean, what a UMI or UMF of 2 or 1.8 or 1.5? I feel like that's not something that at least, as of yet to be great for you, if that became something about which people had an intuition and a natural understanding. It was like a number they were used to kind of thinking about where it made trend and where it may fall.
你怎么帮助放贷人思考，我是说，一个UMI或UMF的2、1.8或1.5是什么意思？我觉得这对你来说至少目前还不是很有好处，如果这成为人们有直觉和自然理解的事情。就像一个数字，他们习惯于考虑它在哪里趋势，并且可能会下降。

But how do you think about like what causes a UMI or UMF to get to a certain level? I mean, I think that scenario planning of, you know, if you think about just the Dodd-Frank stress test kind of thing, if unemployment goes here and something goes here like how bad does that get? What will have my portfolio? How do you think about connecting the dots between those concept of what might happen and what that will mean to where these loans could be performing?
你怎么看待UMI或UMF达到某个水平的原因呢？我的意思是，如果你想象一下多德-弗兰克压力测试的场景规划，如果失业率上升、某些事情变糟，这会有多严重？我的投资组合会受到什么影响？你会怎样把这些可能发生的事情联系起来，看看这对这些贷款的表现意味着什么？

Yeah, I think there are two good ways to build intuition around these UMI index values. The first is to go back to that personal savings rate, which is largely correlated, not perfect, but pretty good. And again, it includes the unemployment and inflation, all the pieces. And what you can basically see is that you can think of it as a historically average level of the personal savings rate, let's just be round numbers, call it on the order of 10% roughly corresponds to a sort of normal UMI condition of around 1.0.
嗯，我认为有两种好的方法来建立对这些UMI指数值的直觉。第一种方法是回到那个个人储蓄率，它与UMI指数存在着很大的相关性。虽然没有完美，但还是比较好的。而且它包括失业率和通胀率等各个方面。你可以把它想象成个人储蓄率的历史平均水平，就称它为大约10%的水平，这大约对应着UMI指数大约为1.0的正常情况。

And then, you know, what happens, of course, if the personal savings rate goes up, you know, if the personal savings rate goes up to something like 20%, which, you know, basically never happens except in 21 when people were given what's up in the market. Yeah, when there's free money going up to go, are people saved on that? Yeah, exactly. But when you get something like that, then you see UMI basically get cut in half down to about 0.5. And again, I'm just speaking in round numbers here.
然后，你知道的，如果个人储蓄率上升，当然会发生什么，如果个人储蓄率上升到20%左右，这基本上几乎从来不会发生，除非在21年当人们得到了市场上的利润。是的，当有免费的钱可以拿去存储时，人们会存储的。是的，没错。但当你得到这样的东西时，那么你会看到UMI基本上被削减了一半，降至0.5左右。再次强调，我只是使用大致的数字。

And then on the other hand, you can see that UMI values go from 1.0 to 2.0 if you basically roughly speaking, cut your personal savings rate in half, which is on the order of what we've seen over the past 18 months or so. And then, the other sort of way to build intuition is just to look at different periods of time. So again, UMI 1.0 roughly corresponds to the five years before the COVID period in the sort of happy days of free flowing stimulus money.
另一方面，你可以看到，如果粗略地将个人储蓄率减半，UMI数值就会从1.0上升到2.0，在过去的18个月左右，我们已经见证了这种减半的情况。另外，建立直觉的另一种方法就是看不同时间段。因此，UMI 1.0大致相当于COVID前五年的那段时期，也就是那个洪流般的刺激资金充沛的幸福时光。

You know, you've got UMI's that go down to nearly 0.5. And then, you know, stimulus running out, producing that large shock takes you UMI over 1.1.1.1.5. And then, if you were to sort of look at historical values of personal savings in UMI, sort of ask the question, well, how bad can it get? At least, how bad have we seen it get? Historically, probably the answer is something like, you know, in the worst times ever, that, again, in recent recording history, we have all the relevant economic data, maybe something like the worst month of the Great Recession we think UMI would have been in the high twos.
你知道，你的UMI可以下降到接近0.5。然后，如果刺激效应消失了，产生了大冲击，你的UMI可能会超过1.1.1.5。如果你去看UMI个人储蓄的历史值，你可能会想知道，情况会有多糟糕？至少，我们有哪些记录表明了最糟糕的情况？从历史上来看，可能是在最糟糕的时代，比如最近经济数据都有的大衰退的最糟糕的一个月，我们认为UMI可能会高达两位数。

Interesting. I get two more things I want to hit on, and I think we've kind of, at least gotten a good intro for people into this concept. One is, we didn't really talk about this, but UMI, as you've built it today, is really based explicitly on unsecured installment loans, particularly the kind that upstart partners with banks and credit unions to originate. Are there intentions to build something like UMI for other asset classes? Is it applicable to other asset? Like, how do you think about the usefulness of it as a construct or the limitations of it as a construct given that it's really a measurement specifically on a particular category of consumer credit?
有趣。我还想谈论两个问题，我认为我们至少已经为人们介绍了这个概念的好入门。其中一个是，我们没有真正谈论过这个，但是像您今天建立的UMI，实际上是明确基于未担保的分期付款贷款，特别是upstart与银行和信用社合作发起的那种。是否有意愿为其他资产类别构建类似于UMI的东西？它是否适用于其他资产？比如，您如何思考它作为一种构建的实用性或限制性，考虑到它实际上是一个特定的消费信贷类别的度量？

Yeah. Great question. So, we don't see UMI as something that you just develop methodology once and it's done. It would be nice if it was like that, but really, it's a very sort of rich surface to explore, because there's not only the question you mentioned, which is the question of, you and my for different products, but there's actually, you and my for different types of consumers, and while maybe on average, they will all move together. They don't, of course, exactly all move together. Their slopes are a little bit different. Their sensitivities are a little different.
是的，很好的问题。我们不认为UMI只需要开发一种方法，然后就完成了。如果能这样就太好了，但实际上，UMI是一个非常丰富的领域需要探索，因为不仅有你和我的不同产品的问题，而且实际上，还有不同类型消费者的你和我，虽然他们可能平均一起移动，但当然并非完全一致。他们的斜率有些不同，敏感度也略有不同。

And so, what we really are, you can think of the progression of UMI methodologies as going from first, you sort of assume there is just one UMI, quote, quote, working off one reference model. Then you have one UMI that's produced actually by an amalgamation of reference models, and that sort of gets more and more continuous as you go from one to two to N, and then you really get to a methodology that's sort of much more continuous with respect to your reference model.
所以，真正的意思是，你可以把UMI方法的发展看作是从最初假设只有一个UMI，基于一个参考模型开始，发展成一个由不同参考模型融合而成的UMI，随着数量从一个到两个再到N不断增加，这一过程会变得越来越连续，最终你会得到一种更加连续的基于你的参考模型的方法。

And then you have a sort of second dimension of UMI methodology improvement, which is really about going from assuming sort of quote, quote, the average bar award to really making it something that were the UMI is interacting with the borrower characteristics, not unlike how we see all sort of the other work we do in sort of prediction realm, is there's just interaction effects going on between almost any given characteristics of a borrower, and of course, those also interact with macron, so those should be considered.
然后，你还有一种UMI方法改进的第二维度，这实际上是从假设一般的引用条款奖项到真正使其成为UMI与借款人特征相互作用的东西，这与我们在预测领域中进行的各种其他工作非常相似，几乎每个借款人的特征都存在交互作用, 当然，它们也与Macron相互作用，因此应该考虑到这些。

And then the last sort of dimension is the type of loan that it is, and that's true of its size, of its duration, of its type, and so that could be it being an auto loan, a personal loan, a three-year loan, a one-year loan, or any sort of combination of these. And so there's a lot of sort of continuing research into UMI that we're doing.
然后，最后一种维度是贷款的类型，包括贷款的规模、期限和种类，可能是汽车贷款、个人贷款、三年贷款、一年贷款，或者这些的任何组合。因此，我们正在进行大量关于UMI的持续研究。

And having said all that, the good news is that we think even this first version of UMI as a first approximation, if you just want to answer the question of under what conditions will my loans over underperform? And by roughly how much, you're going to get a very good approximation just from UMI V1. That is, it's true. It's built on unscured personal loans, but we look at it in the context of our other products like our auto-refi products.
说了这么多，好消息是我们认为即使是UMI V1的第一个版本，如果您只是想回答“我的贷款在什么情况下会表现不佳？”以及“大约会差多少？”这个问题，您仍然能从UMI V1中获得非常好的近似值。也就是说，这是真的。它是建立在未担保的个人贷款上的，但我们会将它与我们的其他产品，比如我们的车贷再融资产品放在一起考虑。

And again, it's not quite as precise, but it does give you a really close first order approximation of your level of over-underperformance, again, looking at those UMI, UMF ratios of the changes across time. It does a good job explaining, and that gives us more confidence in the methodology because it tells us that there's something very general happening in addition to the sort of many smaller but more specific things going on within each product, within each type of borrower and so on and so forth.
再说一遍，这并不是非常精确，但它确实可以给你一个非常接近于你的过度或不足表现水平的近似值，再次观察那些在时间上变化的UMI、UMF比率。它做了一个好工作，给我们更多的方法来解释方法，因为它告诉我们不仅存在于每种产品、每种借款人等等许多更具体的事情中，而且还有一些非常普遍的东西正在发生。

Interesting. I guess my last question for you is, it sounds like this is early days in the development of UMI, like a very solid foundation, but compared to the work you put into understanding microarrisc, it's still pretty immature in comparison and has a lot of room for improvement.
有意思。我猜我最后一个问题是，听起来UMI的开发还处于早期阶段，像非常扎实的基础，但与你投入理解microarrisc的工作相比，它仍然相对不太成熟，并有很大的改进空间。

And you also mentioned if you want to get a new tuition, they should look at the numbers. So if people are interested in looking at the numbers or following along as a methodology and kind of the insights coming out of the work into macro and that UMI evolved, is there a place they can go to see those?
你还提到，如果想申请新的学费，他们应该看看数字。那么，如果人们对 macro（宏观）和 UMI 的工作中涌现的方法和见解感兴趣，有没有一个可以看到这些的地方？

Are they published somewhere? Is there a place they can follow along? What's happening? Where could people go to learn more and keep up the date with what's happening to UMI, what's happening to the work and looking into this macro concept?
它们有没有被发布在某个地方？有没有可以跟进的地方？发生了什么？人们应该去哪里了解更多，并跟上UMI的最新动态，关注这个宏观的概念？

Upstart.com/slash UMI, just out and we're really excited about it and it's going to get better and better but it's going to have the core data there and there's going to be more ways to engage with and play with it and really understand the numbers. So yeah, go there, check it out, see how those values have changed across time, what they were in different periods of time and I think that will really build a good level intuition for this and certainly it's our hope that this would become a tool of general interest beyond those who are either making loans through our platform or even making loans in general but really we think it's a really good and useful tool to just understand what's going on in the world.
Upstart.com/slash UMI 刚出来了，我们很兴奋，它会变得越来越好，但它会有核心数据，在与之交互和玩耍的方式上，还会有更多方法，真正理解这些数字。所以，去那里看看，看看这些价值在不同时间段内如何变化，我认为这将真正建立良好的直觉，我们肯定希望这将成为一个广泛感兴趣的工具，超越了那些通过我们的平台做贷款，甚至是一般贷款人，但我们真的认为这是一个非常好的有用的工具，只是理解世界发生了什么。

Upstart.com/slash UMI is not a very original URL, very straightforward and easy to remember. So I guess that's probably better than original.
Upstart.com/slash UMI 这个网址并不是非常原创，但很直接简单易记。所以我想这可能比原创更好啦。

And I look forward to seeing that. I'm actually really excited that there's, we're going to be publishing the data set so kind of a historical data set. I think there's a lot of rich surface area for innovation on top of it. So I'll be excited to see and if any of the listeners have questions or feedback on what's useful or what you'd like to learn more about the concept, please do write to us or write to just me at Jeff at Upstart.com because I think it's actually a really interesting space and would love to have real input from all of you in terms of where we take the research and what we learn next.
我很期待看到那个。我真的很兴奋，因为我们将会发布这个数据集，是一个历史数据集。我认为它有很多丰富的创新表面。所以我很期待看到它，如果任何听众有关于这个概念有用或者想更多了解的问题或反馈，请写信给我们或写信给我 Jeff，邮箱是 Upstart.com，因为我认为这是一个非常有趣的领域，很愿意听到大家的真实意见，看看我们将会研究何处并且接下来学到什么。

All right Paul, thanks for taking the time and joy. So this is I think a great first dive and I appreciate your coming back for your second time on the podcast.
好的保罗，谢谢你抽出时间和精力。我认为这是一次很好的首次尝试，你能回到播客节目参加第二次我也很感激。

Awesome, great. Thanks for having me. Upstart partners with banks and credit unions to help grow their consumer loan portfolios and deliver a modern all digital lending experience. As the average consumer becomes more digitally savvy, it only makes sense that their bank does too.
太棒了！非常感谢您邀请我来。Upstart 与银行和信用社合作，帮助其发展消费者贷款组合，并提供现代化的全数字借贷体验。随着普通消费者变得越来越精通数字技术，同样也只有他们的银行才能跟上。

Upstart's AI lending platform uses sophisticated machine learning models to more accurately identify risk and approve more applicants than traditional credit models with fraud rates near zero. Upstart's all digital experience reduces manual processing for banks and offers a simple and convenient experience for consumers.
Upstart的人工智能借贷平台采用先进的机器学习模型，比传统信用模型更准确地识别风险，并批准更多的申请人，诈骗率几乎为零。 Upstart的全数字体验可减少银行的手动处理，并为消费者提供简单便利的体验。

Whether you're looking to grow and enhance your existing personal and auto lending programs or you're just getting started, Upstart can help.
无论您是想要增强和扩展您现有的个人和汽车贷款项目，还是刚开始，Upstart都能提供帮助。

Upstart offers an end-to-end solution that can help you find more credit worthy borrowers within your risk profile with all digital underwriting, onboarding, loan closing and servicing.
Upstart提供了一套从头到尾的解决方案，可以帮助您在您的风险指标范围内找到更多有信用价值的借款人，具备全数字化的核保、入职、贷款结算和服务。

It's all possible with Upstart in your corner. Learn more about finding new borrowers, enhancing your credit decision process and growing your business by visiting Upstart.com slash 4-Banks.
Upstart在你身边，一切皆有可能。通过访问Upstart.com/4-Banks，了解更多关于寻找新借款人，优化信用决策流程和发展业务的信息。

That's Upstart.com slash 4-Banks.
那是Upstart.com斜杠4-Banks。

You've been listening to Leaders in Lending from Upstart. Make sure you never miss an episode. Subscribe to Leaders in Lending in your favorite podcast player.
你一直在聆听Upstart的领先放贷者节目。一定要不错过任何一集，订阅你最喜欢的播客播放器中的领先放贷者节目。

Using Apple Podcasts, leave us a quick rating by tapping the number of stars you think the show deserves.
用苹果播客，点击你认为该节目值得的星星数量，给我们留下一个快速评价。

Thanks for listening. Until next time.
谢谢你们的聆听。下次再见。