Transform risk analytics into Competitive edge for Capital Management

45% lower Economic Capital requirement on a typical corporate loan portfolio ? YES, it is possible with ALGOSAVE. And not just during the Tour de France.

45% lower economic capital requirements on a typical corporate loan portfolio ? yes it is possible with ALGOSAVE proprietary financial technology. And not just during the Tour de France.

45% lower economic capital requirements on a typical corporate loan portfolio ? yes it is possible with ALGOSAVE Financial Technology combination of
– borrower & seniority specific Loss Given Default (LGD) term-structure, and
– deep financial-analysis driven Default Probability (PD) and Joint-PD correlation term-structure.

Background :

The main drivers of Economic Capital estimates are :
– Loss Given Default (LGD),
– default probability (PD),
– and at a credit portfolio level, Joint PD correlation.

Let’s review each of those critical component with ALGOSAVE proprietary Financial Technology and see why and how ALGOSAVE helps banks save substantial capital.

1 – Seniority specific and economic-cycle sensitive Point in Time Loss Given Default (LGD) term structure:

ALGOSAVE addresses the 2 main drawbacks of traditional LGD “Beta” distribution model :
– uni-modality : where empirical evidence show that recoveries are – at least – bi-modal, with typical peaks at 20% and 80%
– difficulty to cope with “point masses” at 0% and 100%, where empirical evidence also show that LGD are often located in those areas.

Simple solution : ALGOSAVE financial technology GOT RID of LGD Beta model.

ALGOSAVE proprietary Financial Technology (FinTech) projects corporate financial statements – 100 data points – using Montecarlo simulations. This allows ALGOSAVE FinTech to reflect each corporate specific asset and leverage structure in each of its thousands of Montecarlo simulations paths at 3 different level of seniority : secured, unsecured and subordinated.

For instance, let’s assume that on a given path of its Montecarlo projections, McDonald (MCD) net debt is greater than its Enterprise Value in year 1. This comes – for instance – from higher capital expenditure than average. Assuming that, at the same time, MCD must draw on its Revolver Credit Facility – for instance to fund a greater than optimal sales growth rate – ALGOSAVE Financial Technology “pushes” MCD in default on that specific path.

ALGOSAVE then computes recovery using MCD specific debt and MCD specific asset liquidation value upon default . The latter takes into account 2 possible Machine Learning based outcomes : (1) a gone concern liquidation value, where MCD is taken over (at a multiple of sales or EBITDA) and keeps operating – or (2) a fire sale of MCD assets which price depends on the then Machine Learning Based economic cycle.

Following this methodology, this graph depicts the unsecured 1-year LGD distribution of a typical portfolio of 100 corporates selected from ALGOSAVE issuer database. And not Singapore skyline. The x-axis measures the LGD between 0 and 1.0 while the y-axis measures the number of observations of each bar in the histogram

First observation ? Both the multi-tail and “point mass” nature of ALGOSAVE proprietary LGD distribution clearly shine.

The blue LGD histogram depicts the LGD distribution for OECD consensus growth scenario with 59% average LGD – close to Basel Foundation 55% LGD and CDS 60% standard ISDA LGD. Examining the red histogram – which depicts the LGD distribution for ALGOSAVE stressed macro economic scenario – it is interesting to notice the marked shift of the LGD distribution from left (15% “blue” LGD) to right (100% “red” LGD). The average ALGOSAVE stressed-scenario LGD stands at 72%.

2 – Point in Time Probability of Default term structure :

ALGOSAVE Probability of Default (PD) term structure is calibrated on capital market PD term structure which comes from CDS, bonds and/or secondary loan trading.
It can also be calibrated on ALGOSAVE clients’ internal Point-in-Time default probability term structure.

This calibration gives ALGOSAVE PD term structure its unique “live” flavor which – by the way – is required by new accounting regulations.

3 – Joint PROBABILITY OF DEFAULT correlation :

ALGOSAVE proprietary joint PD correlation technology is a lot lower than its ill-fated CDS and equiy price correlation proxies. Those are full of “market noise” such as iTraxx CDS sector and ETF trading and hide real default correlation.

But, CLO traders and loan/bond portfolio managers are well aware of this phenomenon whereby actual default correlation – to the exception of those in the highly regulated financial sector – is a lot lower than its capital market proxies.

You may be interested by one of our latest posts which dealt with this issue in greater detail. Read it all here.

Conclusion – ALGOSAVE proprietary Financial Technology helps banks save 45% in economic capital

Indeed, putting all things together – LGD, PD and Joint PD correlation – where traditional Expected Loss model points toward a 3.30% Economic Capital requirement (on the sample corporate portfolio), ALGOSAVE proprietary Financial Technology only requires 1.80% Economic Capital at 99.9% confidence level.

A 45% saving on Economic Capital !

Now, that’s something worth celebrating. And, not just among our Welsh Tour de France cycling fans 🙂

Transform risk analytics into Competitive edge for Capital Management

Good century-old DuPont-analysis helps Banks SAVE CAPITAL : what an explosive surprise !

 

Good century-old DuPont-analysis helps Banks SAVE CAPITAL : what an explosive surprise !

ALGOSAVE confirms DuPont analysis profound wisdom : LGD CORRELATION is ROCK-BOTTOM LOW and helps financial institutions SAVE CAPITAL.

And, especially for high grade borrowers, ALGOSAVE has another BIG surprise in store on PD-LGD correlations.

DuPont Analysis comes from the DuPont Corporation that started using this formula in the 1920s.
DuPont explosives salesman Donaldson Brown invented this formula in an internal efficiency report in 1912.

More than a century after that, ALGOSAVE CORPORATE DATABASE shows that DuPont analysis also hides another even more profound wisdom : the CORRELATION of Loss Given Default (LGD) between corporates in a given industry is ROCK BOTTOM LOW to negative. This means IMMEDIATE CAPITAL SAVING for corporate lenders.

In The Beginning, long time ago.

DuPont analysis tells us that the Return On Equity – ROE – can be decomposed in 3 items

Net Income / Sales = Profitability
Sales / Total Assets = Asset Efficiency
Total Assets / Average Shareholder Equity = Financial Leverage

  1. DuPont analysis also tells us that profitability is mostly matter of technology.
    In other word, when a company is part of a given industry, it “inherits” the industry profitability which is technology-dependent.
    For instance, if you distribute food in the US, your expected Net Income / Sales cannot be too-far away from your peers. Unless you have a completely different service or…technology.
  2. DuPont analysis also tells us that Asset Efficiency is mostly a matter of the corporate competitive landscape.
    Indeed, in order to increase this ratio, the company will have to gain market share <=> increase marketing expenses, lower price, increase inventories to prepare for increased sales, and give longer credit to its clients <=> lower net Income and higher Total Assets.
    Profitability and Asset Efficiency are interdependent.
  3. So, in order to increase its ROE and make a difference, the company’s management is mostly left with the latest ROE key driver : ITS degree of Financial Leverage. Let’s also keep in mind that an increase in financial leverage also means an increase in the Equity BETA, with its consequence on WACC and ultimately on the Expected ROE.


From DuPont analysis … to rock-bottom LOW LGD CORRELATION

If DuPont analysis holds true – and indeed corporates in the same industry mostly drive their ROE thru financial leverage – then when they go bankrupt together, we should be expecting little correlation between their respective Loss Given Default : they all finance their assets in a different way to make a difference in their respective ROE.

The challenge : historical high-grade corporates LGDs are scarce <=> Default of large and solid corporates are a rare event.
Concomitant defaults of such corporates are even rarer. So that measuring historical LGD correlation is a challenge.
Estimating forward looking and Point in Time LGD correlation is a double-challenge.

Algosave technology raises to the challenge and delivers its clients exactly that : forward looking and Point in Time stress-tested LGD correlations.
ALGOSAVE delivers those for all the 5000 borrowers in ALGOSAVE CORPORATE DATABASE.

Let’s take an example on our favorite Integrated Oil and Gas basket : 11 high-grade global corporates

ALGOSAVE CORPORATE DATABASE delivers the following data : average 5-year unsecured stressedLGD, current and stressed cumulative 5 year PD

  1. Before the bigger surprise, let’s have a look at 5-year average stressed unsecured LGDs for each of those 11 corporates.Although the average LGD of those asset-heavy corporates is close to 49% – which is close to CDS standard LGD  – there is a marked difference between BP P.L.C. 81% 5-year LGD and Husky Energy 22% 5-year LGD.
    This already confirms DuPont analysis differing financial structure, albeit in a static and average way.
  2. Let’s also compare current market 5-year PDs and ALGOSAVE stressed 5-year PDs.
    Although current 5-year cumulated PDs are multiplied by a factor of 5 when stressed (increase from 4% to 20%), BP and Statoil are multiplied by a factor of more than 10, whereas Husky Energy is only multiplied by a factor of 2.5.
    This also confirms DuPont analysis about differing financial structure, albeit in a static and average way.
  3. Finally, last but not least, ALGOSAVE CORPORATE DATABASE unique and CAPITAL SAVING deliverable : Forward-looking, Point in Time and stressed 5 year unsecured LGD correlations

    On average stressed LGD correlation is a ROCK BOTTOM -0.01
    The highest LGD correlation is a small 0.11, between Royal Dutch Shell and EXXON.
    The lowest LGD correlation is also a small -0.13, between TOTAL SA and Suncor Energy.
    Thinking “DuPont”, this should not be surprising. Indeed, DuPont analysis implicitly states that in order to drive its ROE, company management is mostly left with carefully choosing its Degree of Financial Leverage. Hence is case of concomitant default of two corporates, lenders should not expect to loose the same amount of money on their debt.As a conclusion, ALGOSAVE confirms DuPont analysis profound wisdom : LGD CORRELATION is a ROCK-BOTTOM LOW and CAPITAL SAVING critical metric.
  4. Last but not least ALGOSAVE ALSO offers a surprise on PD-LGD correlation in stressed scenario for high-grade borrowers.
    This will be the object our our next post. Please stay tuned.If you like this post, do not hesitate to ask for you free subscription :